Luisa PetersonIon Victor GoseaPeter BennerKai Sundmacher...
108917.1-108917.21页
查看更多>>摘要：A digital twin (DT) is an automation strategy that combines a physical plant with an adaptive real-time simulation environment, where both are connected by bidirectional communication. In process engineering, DTs promise real-time monitoring, prediction of future conditions, predictive maintenance, process optimization, and control. However, the full implementation of DTs often fails in reality. To address this issue, we first examine various definitions of DTs and its core components, followed by a review of its current applications in process engineering. We then turn to the computational and numerical challenges for building the simulation environments necessary for successful DTs implementation.
Luis Antonio Pulido-VictoriaAntonio Flores-TlacuahuacAlexander Panales-PerezTania E. Lara-Ceniceros...
108920.1-108920.13页
查看更多>>摘要：Ceramic 3D printing has become an increasingly popular manufacturing technique due to its potential to create complex geometries with high precision. However, predicting the printability of ceramic pastes remains a challenge, as it depends on various rheological properties. In this study, we propose a feed-forward deep neural network model that predicts the printability of ceramic pastes based on two suggested criteria, a shear-thinning ability and a gel point greater than 1×10~3 Pa. The model is trained on a dataset built from rheological and viscoelastic characterizations of pastes, and validated on a separate test set. Our results show that the proposed learning model achieves small relative error in predicting the gel point of the ceramic pastes, with a mean value of 8.99181 and a standard deviation of 1.812864. This model has the potential to improve the efficiency and quality of ceramic 3D printing by enabling rapid and accurate predictions of paste printability.
查看更多>>摘要：Frequently, parameters in optimization models are subject to a high level of uncertainty coming from several sources and, as such, assuming them to be deterministic can lead to solutions that are infeasible in practice. Robust optimization is a computationally efficient approach that generates solutions that are feasible for realizations of uncertain parameters near the nominal value. This paper develops a data-driven robust optimization approach for the scheduling of a straight pipeline connecting a single refinery with multiple distribution centers, considering uncertainty in the injection rate. For that, we apply support vector clustering to learn an uncertainty set for the robust version of the deterministic model. We compare the performance of our proposed robust model against one utilizing a standard robust optimization approach and conclude that data-driven robust solutions are less conservative.
查看更多>>摘要：In industrial processes, limitations of the physical environment, sensors drop-out, and repetitive sampling often lead to insufficient and unevenly distributed representative instances, which greatly hinders the accuracy of soft-sensing models. This paper presents a novel virtual sample generation method based on Glow-embedded variational autoencoder (GVAE-VSG), aimed at enhancing data richness and diversity to improve the modeling performance. Specifically, GVAE-VSG embeds the Glow model from flow transformations into the variational autoencoder. This allows for the derivation of a more generalized posterior distribution without reducing sample dimensionality, thereby ensuring the generation of higher-quality virtual input samples. Subsequently, a nonlinear iterative partial least squares regression framework, incorporating a sparse constrained error matrix, is employed to generate virtual output samples that more closely resemble actual data. Finally, by a synthetic nonlinear function and an actual purification terephthalic acid (PTA) solvent system, the generative and modeling performance of the proposed method are comprehensively assessed.
查看更多>>摘要：Growing energy demand and its consequences, such as fossil fuel depletion, greenhouse gas emissions, and global warming, prompted the need for large-scale solar power plants. Floating photovoltaic systems have many advantages over ground-mounted systems, including methods and resources, reducing costs, and improving efficiency. In this regard, this study aims at presenting an optimization model for developing a sustainable and resilient floating solar photovoltaic supply chain network design. The concerned model's objective function is minimizing the total supply chain costs in addition to maximizing greenhouse gas emissions reduction. To identify the most suitable dams for establishing the floating photovoltaic system, the hybrid approach by applying the fuzzy best-worst method and the TOPSIS technique is first exploited. Thereinafter, the selected dams are exerted in the presented mathematical model. Eventually, a real case study is implemented on floating photovoltaic systems to assess the proposed model's performance, from which important managerial insights are attained.
查看更多>>摘要：This paper presents a methodology of hyperparameter optimization (HPO) of deep neural networks for molecular property prediction (MPP). Most prior applications of deep learning to MPP have paid only limited attention to HPO, thus resulting in suboptimal values of predicted properties. To improve the efficiency and accuracy of deep learning models for MPP, we must optimize as many hyperparameters as possible and choose a software platform to enable the parallel execution of HPO. We compare the random search, Bayesian optimization, and hyperband algorithms, together with the Bayesian-hyperband combination within the software packages of KerasTuner and Optuna for HPO. We conclude that the hyperband algorithm, which has not been used in previous MPP studies, is most computationally efficient; it gives MPP results that are optimal or nearly optimal in terms of prediction accuracy. Based on our case studies, we recommend the use of the Python library KerasTuner for HPO.
查看更多>>摘要：Accurate prediction of lithium-ion battery lifespan is crucial for mitigating risks, as battery cycling experiments are time-consuming and costly. Despite this, few studies have effectively leveraged cycling data with minimal information loss and optimized input size. To bridge this gap, we propose three models that integrate attention layers into a foundational model. Temporal attention helps address the vanishing gradient problem inherent in recurrent neural networks, enabling a manageable input size for subsequent networks. Self-attention applied to context vectors, termed cyclic attention, allows models to efficiently identify key cycles that capture the majority of information across cycles, strategically reducing input size. By employing multi-head attention, required input size is reduced from 100 to 30 cycles, significant reduction than single-head approaches, as each head accentuates distinct key cycles from various perspectives. Our enhanced model shows a 39.6 % improvement in regression performance using only the first 30 cycles, significantly advancing our previous method.
查看更多>>摘要：This work introduces a semi-supervised learning algorithm to estimate missing data for processes where measured data is comprised of variables that are measured at high frequency and low frequency. A semi-supervised learning algorithm named "Weight-Adjusted Consistency Regularization Algorithm for Semi-Supervised Learning" (WACR-SSL) based on consistency regularization is proposed. The algorithm splits the irregular unbalanced data set into three parts and processes them separately. To address the loss balancing problem, five loss balancing methods have been tested: Uncertainty Weights (UW), Random Loss Weighting (RLW), Dynamic Weight Average (DWA), Geometric Loss Strategy (GLS) and the logarithmic transformation (LogT). When applied to data from a hydrocracking process, the algorithm effectively leverages partially labeled data. With carefully chosen noise scales and the coefficient for the unsupervised loss, the uncertainty weight (UW) variant performs the best when compared to the other loss balancing methods.
Ali GhodbaAnne RichelleChris McCreadyLuis Ricardez-Sandoval...
108935.1-108935.9页
查看更多>>摘要：The study proposes a robust algorithm for batch-to-batch optimization in the presence of model-mismatch. Robustness is achieved by the implementation of the following features: ⅰ - the gradient correction step is modified to consider the gradients of the cost function and constraints at both final and intermediate points, ⅱ - Economic Model Predictive Control is applied to mitigate the impact of unmeasured disturbances on the optimum, and ⅲ - an optimal design of experiments is performed to expedite convergence. Significant improvements of the proposed algorithm in convergence to the process optimum and robustness to noise, unmeasured disturbances, and model error are demonstrated using a fed-batch fermentation for penicillin production.
查看更多>>摘要：Kinetic Monte Carlo (kMC) models are a well-established modelling framework for the simulation of complex free-radical kinetic systems. kMC models offer the advantage of discretely monitoring every chain sequence in the system, providing full accounting of the chain molecular weight distribution. These models are marred by the necessity to simulate a minimum number of molecules, which confers significant computational burden. This paper adapts and creates a highly generalizable methodology for scaling dilute radical populations in discrete stochastic models, such as Gillespie's Stochastic Simulation Algorithm (SSA). The methodology is then applied to a kMC simulation of polystyrene CPS) pyrolysis, using a modelling framework adapted from literature. The results show that the required number of simulated molecules can be successfully reduced by up to three orders of magnitude with minimal loss of convergent behaviour, corresponding to a wall-clock simulation speed reduction of between 95.2 to 99.6 % at common pyrolysis temperatures.
NETL
NSTL
Elsevier