Article
Chemistry, Analytical
Yuxi Li, Gang Hao
Summary: This paper proposes an improved modified model predictive control algorithm by combining the Sage-Husa adaptive Kalman filter (SHAKF), the cubature Kalman filter (CKF), and the back-propagation neural network (BPNN) to mitigate the negative impacts of system noise on energy-optimal adaptive cruise control (EACC) and achieve further energy reduction.
Article
Thermodynamics
Nitesh Kumar Sahu, Mayank Kumar, Anupam Dewan
Summary: A computational model was developed to study the impact of three operating parameters on gasifier flow field and coal gasification process, showing that their interaction governs optimal operation.
Article
Engineering, Electrical & Electronic
Fatih Cengil, Harsha Nagarajan, Russell Bent, Sandra Eksioglu, Burak Eksioglu
Summary: Machine learning-based methods are proposed to accelerate convergence to global solutions for the AC Optimal Power Flow problem. By leveraging historical data, a subset of variables can be selected to tighten bounds and find near-global optimal solutions at faster run-times.
ELECTRIC POWER SYSTEMS RESEARCH
(2022)
Article
Engineering, Marine
Lifen Hu, Ming Zhang, Zhi-Ming Yuan, Hongxia Zheng, Wenbin Lv
Summary: As offshore engineering moves into deeper waters, floating structures have become crucial in offshore structure communities. This study proposes a predictive control strategy based on a machine learning algorithm to minimize the heave motion of crane payload. Experimental results demonstrate that the proposed control strategy significantly reduces the vertical motion of the payload.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2023)
Article
Automation & Control Systems
Qibo Liu, Shaoyuan Li, Yi Zheng, Chenkun Qi, Min Luo
Summary: This study presents a learning-based approximation scheme to reduce the computational burden of distributed model predictive control (DMPC). An independent neural network approximator is designed for each subsystem to ensure the feasibility and stability of the global system. Simulation results demonstrate the effectiveness and superior performance of the proposed strategy.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Energy & Fuels
Yuchong Huo, Zaiyu Chen, Jing Bu, Minghui Yin
Summary: This study proposes a column generation approach assisted by deep neural networks to solve the model predictive control problem, which can accelerate the computation process, reduce computational cost, and ensure the feasibility and optimality of the solutions.
Article
Green & Sustainable Science & Technology
Youngtak Cho, Gyuyeong Hwang, Dela Quarme Gbadago, Sungwon Hwang
Summary: In this study, an artificial neural network (ANN) model was developed and integrated with model predictive control (MPC) to provide optimal conditions for operating the PEMFC system. The developed NNMPC yielded improved system power by maintaining the optimal stack temperature, cathode pressure and membrane hydration for current changes at a low computational cost. The optimized power of the PEMFC system showed an average increase of 10.9% compared to the fixed-setpoint condition.
JOURNAL OF CLEANER PRODUCTION
(2022)
Article
Construction & Building Technology
Lei Lei, Wei Liu
Summary: This study investigates the difficulties in room temperature control in a multi-zone variable air volume (VAV) air-conditioning system and proposes a predictive control model based on radial basis function (RBF) neural network. Experimental results show that the proposed model can meet room temperature requirements and ensure stable static pressure of the main air supply duct.
ENERGY AND BUILDINGS
(2022)
Article
Energy & Fuels
Sonja Kallio, Monica Siroux
Summary: In a solar micro-grid, a hybrid renewable energy system is used to generate electricity for onsite use. Accurate prediction of photovoltaic (PV) panel output is important for optimal energy flow control and is challenging due to the fluctuating nature of solar radiation availability. This study compares the accuracy of prediction models using multiple linear regression (MLR) and artificial neural network (ANN) methods, with the ANN model using individual PV output data showing the highest accuracy. The use of micro-inverter technology improves the accuracy of PV prediction for control purposes.
Article
Energy & Fuels
Shiyu Yang, Man Pun Wan, Wanyu Chen, Bing Feng Ng, Swapnil Dubey
Summary: This study proposes an approximate MPC method that mimics the dynamic behaviors of MPC using recurrent neural networks, with implementation in two testbeds showing similar control performance to MPC and significantly reduced computational load.
Article
Automation & Control Systems
C. Courtes, E. Franck, K. Lutz, L. Navoret, Y. Privat
Summary: Using individual-based models instead of classical population-based models can overcome their shortcomings by considering heterogeneity features and describing small cluster dynamics. However, these models often involve large graphs that are costly and difficult to optimize. This study proposes a numerical approach combining reinforcement learning philosophy with reduced models to determine optimal health policies for stochastic individual-based models with heterogeneity. The approach involves a deterministic reduced population-based model with a neural network that mimics the local dynamics of the individual-based model. The optimal control is determined by sequentially training the network until the population-based model successfully contains the epidemic in the individual-based model.
OPTIMAL CONTROL APPLICATIONS & METHODS
(2023)
Article
Engineering, Chemical
Esin Iplik, Ioanna Aslanidou, Konstantinos Kyprianidis
Summary: This study proposes a feedforward model predictive control structure for an industrial hydrocracker and compares different models in terms of reactor temperature decisions and diesel product quality predictions. The results highlight the importance of feed character measurements and demonstrate significant improvements in product quality and energy savings.
Article
Green & Sustainable Science & Technology
Lu Wang, JianJuan Yuan, Xu Qiao, Xiangfei Kong
Summary: Solar energy coupled with electric heat storage is a promising energy saving technology for distributed building heating. Precise heat load prediction and renewable energy prediction are necessary for building clean energy supply systems. A predictive control based on BP artificial neural network is used to predict the demand heat load and heat supply quantity of a solar collector system. A double predictive control is introduced to optimize the operation of renewable energy, heat storage and electric heating.
Article
Engineering, Marine
Min-Kyung Lee, Inwon Lee
Summary: The optimal design of flow control fins (FCFs) for a container ship was carried out using a machine learning approach. Instead of simulation-based performance evaluation, artificial neural network (ANN) was used for prediction. Systematic collection of wake distribution data via computational fluid dynamics (CFD) was done before the machine learning process. The results showed that both wake distributions and resistance performance were improved in a practically applicable timeframe.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2023)
Article
Automation & Control Systems
Xing Liu, Lin Qiu, Youtong Fang, Kui Wang, Yongdong Li, Jose Rodriguez
Summary: This letter introduces an improved control method that extends and improves the classical predictive control approach, addressing the limitations of system uncertainties and unknown perturbations. The method uses unsupervised learning technique to learn the control part online, and incorporates a robustifying control term to enhance the robustness. Unlike traditional methods, this approach does not require prior knowledge of model information and weighting factors, making it applicable to various power converter systems.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Engineering, Mechanical
Zhixin Zhan, Hua Li
Summary: In aerospace engineering, a platform is developed for data-driven fatigue life prediction of AM stainless steel 316L, utilizing ML models like ANN, RF, and SVM. The effectiveness of the platform is verified through comparisons with experimental data, and detailed parametric studies using ML models are conducted to investigate significant characteristics.
INTERNATIONAL JOURNAL OF FATIGUE
(2021)
Article
Engineering, Mechanical
Qimin Liu, Xin Ye, Hangyu Wu, Xingyu Zhang
Summary: This study presents a 3D multiphysics model to investigate the motion and deformation of drug-loaded magnetic hydrogel in a moving fluid under a varying magnetic field. The results show that a slower moving magnet, a larger hydrogel radius, and a faster flow velocity can shorten the time for the hydrogel to reach the channel outlet. Additionally, a magnetic targeting system is proposed for transporting the drug-loaded hydrogel to a specific site by controlling the magnet velocity and maximum magnetic field strength.
INTERNATIONAL JOURNAL OF MECHANICAL SCIENCES
(2022)
Article
Materials Science, Multidisciplinary
Li-Ya Liu, Qing-Sheng Yang, Bo-Jun Miao, Xia Liu, Xing-Yu Zhang
Summary: Graphene/aluminum composites are successfully fabricated using selective laser melting, with optimized printing parameters. Anisotropic and elastoplastic characteristics are explored through nanoindentation experiments, while a mesoscopic crystal plastic finite element (CPFE) model is developed to study the strengthening mechanism of graphene reinforcements in aluminum matrix composites. The CPFE model shows great agreement with experimental data.
MECHANICS OF ADVANCED MATERIALS AND STRUCTURES
(2022)
Article
Nanoscience & Nanotechnology
Xiaohui Song, Xingyu Zhang, Zihao Ou, Ya-Qian Zhang, Mufan Li
Summary: This study presents a facile method for controlling the core position of gold nanoparticles in different shells, utilizing AuNPs with varying sizes, shapes, and surface chemistry. Investigation of the wetting of solid shell domains on seed nanoparticles allows for qualitative measurement of nanoparticle surface wettability, providing a platform for the synthesis of anisotropic nanoparticles.
ACS APPLIED NANO MATERIALS
(2021)
Article
Materials Science, Multidisciplinary
Qian Shi, Mohammed Shahrudin Bin Ibrahim, Xingyu Zhang, Youngkyu Hwang, Hokyun Chin, Shengyang Chen, Wen See Tan, Hua Li, Juha Song, Nam-Joon Cho
Summary: This study unraveled the mechanical and morphological variations in the exine capsules of pollen and spore grains, revealing their impact on germination origin. It also provided insights for selecting pollen and spore species for potential biomaterial applications.
APPLIED MATERIALS TODAY
(2022)
Article
Materials Science, Ceramics
Yang Liu, Muyu Liu, Guitao Luo, Hua Li, Hongbo Tan, Qimin Liu
Summary: A multiphysics-multiscale-multidrive model is developed for C3S hydration. The governing equations are formulated with thermo-chemo-electrical coupled fields, and the multiscale computations are achieved from micro-scale to macro-scale. The multidrives, including C3S dissolution and various gradients, are considered for physiochemical reactions. The model can well characterize the time evolution of hydration heat flow, chemical shrinkage, and ionic concentrations, and the effects of water-to-cement ratios and specific surface areas on C3S hydration kinetics are investigated.
CERAMICS INTERNATIONAL
(2023)
Article
Materials Science, Multidisciplinary
Chengxi Chen, Stanley Jian Liang Wong, Srinivasan Raghavan, Hua Li
Summary: A design of experiments (DOE) informed deep learning (DL) model is developed to address the challenges in high throughput data generation in the directed energy deposition (DED) process. The model accurately predicts the cross-section dilution shape and geometrical characteristics of beads in the single-track deposition of stainless steel 316L. The predicted ranges of porosity and hardness are within acceptable limits.
MATERIALS & DESIGN
(2022)
Article
Engineering, Manufacturing
Chong Heng Lim, Hua Li, Manickavasagam Krishnan, Kewei Chen, Junru Li
Summary: This paper presents a novel method for reducing residual stresses in AlSi10Mg fabricated by laser beam powder bed fusion, without compromising its mechanical properties. By using a lower temperature treatment followed by uneven cooling, thermal stresses induced during cooling cancel out the existing residual stresses, while maintaining the compressive strength of the part.
VIRTUAL AND PHYSICAL PROTOTYPING
(2023)
Article
Chemistry, Physical
Jingtian Kang, Hua Li
Summary: In this study, a multiphysics model was developed to quantify the sensitivity of hydrogels to both hydrostatic pressure and temperature in an electrolyte bathing solution. The proposed model was validated by comparing numerical results with experimental data. The influences of initial fixed-charge density, temperature, hydrostatic pressure, and bathing solution concentration on the volume expansion ratio of hydrogels were investigated. Additionally, the concentration of mobile ions and distribution of electric potential within the hydrogel body and bathing solution were predicted.
PHYSICAL CHEMISTRY CHEMICAL PHYSICS
(2023)
Article
Materials Science, Multidisciplinary
Yang Liu, Hua Li, Muyu Liu, Guitao Luo, Hongbo Tan, Qimin Liu
Summary: This study investigated the role of ion diffusion in the heat flow and chemical shrinkage of tricalcium aluminate (C(3)A) hydration using a multi-ionic reactive transport model. The model considered the coupling of ion diffusion, dissolution, and precipitation reactions during C(3)A hydration in the presence of gypsum. The effects of gypsum content, specific surface area (SSA), and curing temperature on ion diffusion, electrical potential, heat flow, and chemical shrinkage were numerically studied.
JOURNAL OF MATERIALS RESEARCH AND TECHNOLOGY-JMR&T
(2023)
Article
Engineering, Biomedical
Yufeng Shou, Ling Liu, Qimin Liu, Zhicheng Le, Khang Leng Lee, Hua Li, Xianlei Li, Dion Zhanyun Koh, Yuwen Wang, Tong Ming Liu, Zheng Yang, Chwee Teck Lim, Christine Cheung, Andy Tay
Summary: Bone marrow-derived mesenchymal stem cells (MSCs) are widely studied due to their regenerative potential and immunomodulatory properties. In this work, a 3D dynamic hydrogel using magneto-stimulation is introduced for direct manufacturing of MSCs. The technology enhances MSC spreading, proliferation, and biofunctions, and can integrate MSC manufacturing to final clinical use, offering a paradigm shift in MSC therapy.
BIOACTIVE MATERIALS
(2023)
Review
Engineering, Mechanical
Zhixin Zhan, Xiaofan He, Dingcheng Tang, Linwei Dang, Ao Li, Qianyu Xia, Filippo Berto, Hua Li
Summary: This study provides a comprehensive overview of recent developments and future trends in fatigue life prediction of additive manufacturing (AM) metals, with a particular emphasis on machine learning (ML) modeling techniques. It summarizes the recent achievements in fatigue characteristics of AM metals, ML-based approaches for fatigue life prediction, and non-ML-based methodologies. The study aims to guide researchers and engineers in accurately and efficiently predicting fatigue life in AM metal components.
FATIGUE & FRACTURE OF ENGINEERING MATERIALS & STRUCTURES
(2023)
Article
Engineering, Multidisciplinary
Jinze Li, Hua Li, Yiwei Lian, Kaiping Yu, Rui Zhao
Summary: The article proposes two novel implicit methods to meet competitive demands without increasing computational costs. These methods achieve stability and accuracy by introducing auxiliary variables. The article further refines the error analysis and compares the superiority of the novel methods with existing algorithms.
INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING
(2023)
Article
Materials Science, Multidisciplinary
Yang Liu, Muyu Liu, Hua Li, Guitao Luo, Hongbo Tan, Qimin Liu
Summary: This study proposed a theoretical model to investigate multi-mineral reactive transport processes and the effect of multi-mineral reactions on cement hydration kinetics shifting from silicate to aluminate dominance. The model calculates the reaction rates, ionic diffusion, and adsorption based on the ionic concentration, and the hydration heat flow based on the reaction rates. The model was validated and used to study the combined effect of multi-mineral reactions on hydration kinetics.
MATERIALS & DESIGN
(2023)
Article
Materials Science, Multidisciplinary
Xiaohui Song, Zihao Ou, Xiaosong Hu, Xingyu Zhang, Ming Lin, Liaoyong Wen, Mufan Li
Summary: Wet chemical etching has been used to prepare porous metal-organic framework (MOF) particles, with the pH value of the etching solution serving as a controllable parameter to vary the types of porous ZIF-8 particles observed. The 3D electron tomography technique was applied to analyze the porous ZIF-8 particles and reveal their structural properties. The study demonstrates the potential of wet chemical etching and 3D tomography technique in understanding and designing functional porous materials.
ACS MATERIALS LETTERS
(2021)
Article
Mathematics, Applied
Junfeng Cao, Ke Chen, Huan Han
Summary: This paper proposes a two-stage image segmentation model based on structure tensor and fractional-order regularization. In the first stage, fractional-order regularization is used to approximate the Hausdorff measure of the MS model. The solution is found using the ADI scheme. In the second stage, thresholding is used for target segmentation. The proposed model demonstrates superior performance compared to state-of-the-art methods.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2024)
Article
Mathematics, Applied
Dylan J. Oliver, Ian W. Turner, Elliot J. Carr
Summary: This paper discusses a projection-based framework for numerical computation of advection-diffusion-reaction (ADR) equations in heterogeneous media with multiple layers or complex geometric structures. By obtaining approximate solutions on a coarse grid and reconstructing solutions on a fine grid, the computational cost is significantly reduced while accurately approximating complex solutions.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2024)
Article
Mathematics, Applied
Nathan V. Roberts, Sean T. Miller, Stephen D. Bond, Eric C. Cyr
Summary: In this study, the time-marching discontinuous Petrov-Galerkin (DPG) method is applied to the Vlasov equation for the first time, using backward Euler for a Vlasov-Poisson discretization. Adaptive mesh refinement is demonstrated on two problems: the two-stream instability problem and a cold diode problem.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2024)
Article
Mathematics, Applied
Yizhi Sun, Zhilin Sun
Summary: This work investigates the convexity of a specific class of positive definite probability measures and demonstrates the preservation of convexity under multiplication and intertwining product. The study reveals that any integrable function on an interval with a polynomial expansion of fast absolute convergence can be decomposed into a pair of positive convex interval probabilities, simplifying the study of interval distributions and discontinuous probabilistic Galerkin schemes.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2024)
Article
Mathematics, Applied
Bhagwan Singh, Komal Jangid, Santwana Mukhopadhyay
Summary: This paper examines the prediction of bending characteristics of nanoscale materials using the Moore-Gibson-Thompson thermoelasticity theory in conjunction with the nonlocal strain gradient theory. The study finds that the stiffness of the materials can be affected by nonlocal and length-scale parameters, and the aspect ratios of the beam structure play a significant role in bending simulations.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2024)
Article
Mathematics, Applied
Guoliang Wang, Bo Zheng, Yueqiang Shang
Summary: This paper presents and analyzes a parallel finite element post-processing algorithm for the simulation of Stokes equations with a nonlinear damping term, which integrates the algorithmic advantages of the two-level approach, the partition of unity method, and the post-processing technique. The algorithm generates a global continuous approximate solution using the partition of unity method and improves the smoothness of the solution by adding an extra coarse grid correction step. It has good parallel performance and is validated through theoretical error estimates and numerical test examples.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2024)
Article
Mathematics, Applied
Hao Xu, Zeng-Qi Wang
Summary: Fluid flow control problems are crucial in industrial applications, and solving the optimal control of Navier-Stokes equations is challenging. By using Oseen's approximation and matrix splitting preconditioners, we can efficiently solve the linear systems and improve convergence.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2024)
Article
Mathematics, Applied
Zhengya Yang, Xuejuan Chen, Yanping Chen, Jing Wang
Summary: This paper focuses on the high-order stable numerical solutions of the time-space fractional diffusion equation. The Fourier spectral method is used for spatial discretization and the Spectral Deferred Correction (SDC) method is used for numerical solutions in time. As a result, a high-precision numerical discretization scheme for solving the fractional diffusion equation is obtained, and the convergence and stability of the scheme are proved. Several numerical examples are presented to demonstrate the effectiveness and feasibility of the proposed numerical scheme.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2024)