Open Access. Powered by Scholars. Published by Universities.®
- Keyword
-
- Algorithms (2)
- Learning systems (2)
- Least squares approximations (2)
- Neural networks (2)
- Recursive functions (2)
-
- Smart Manufacturing (2)
- Activity Recognition (1)
- Backpropagation (1)
- Breakdown Mechanism (1)
- Chemical Equipment (1)
- Cloud Manufacturing (1)
- Computational Fluid Dynamics (1)
- Computational modeling (1)
- Computer vision (1)
- Control Algorithms (1)
- Control Systems, Nonlinear (1)
- Convergence of numerical methods (1)
- Convolutional Neural Networks (CNN) (1)
- Cross-Flow Instabilities (1)
- Crossflow Vortices (1)
- Cutting force identification (1)
- Cycle Consistency (1)
- Deep Learning (1)
- Digital Twin (1)
- Distillation Column Control (1)
- Disturbance Amplitudes (1)
- Feed rate (1)
- Feedforward Neural Networks (1)
- Force measurement (1)
- Forecasting (1)
Articles 1 - 14 of 14
Full-Text Articles in Mechanical Engineering
Jointly-Learnt Networks For Future Action Anticipation Via Self-Knowledge Distillation And Cycle Consistency, Md Moniruzzaman, Zhaozheng Yin, Zhihai He, Ming-Chuan Leu, Ruwen Qin
Jointly-Learnt Networks For Future Action Anticipation Via Self-Knowledge Distillation And Cycle Consistency, Md Moniruzzaman, Zhaozheng Yin, Zhihai He, Ming-Chuan Leu, Ruwen Qin
Mechanical and Aerospace Engineering Faculty Research & Creative Works
Future action anticipation aims to infer future actions from the observation of a small set of past video frames. In this paper, we propose a novel Jointly learnt Action Anticipation Network (J-AAN) via Self-Knowledge Distillation (Self-KD) and cycle consistency for future action anticipation. In contrast to the current state-of-the-art methods which anticipate the future actions either directly or recursively, our proposed J-AAN anticipates the future actions jointly in both direct and recursive ways. However, when dealing with future action anticipation, one important challenge to address is the future's uncertainty since multiple action sequences may come from or be followed by …
Modeling Of Cloud-Based Digital Twins For Smart Manufacturing With Mt Connect, Liwen Hu, Ngoc-Tu Nguyen, Wenjin Tao, Ming-Chuan Leu, Xiaoqing Frank Liu, Rakib Shahriar, S M Nahian Al Sunny
Modeling Of Cloud-Based Digital Twins For Smart Manufacturing With Mt Connect, Liwen Hu, Ngoc-Tu Nguyen, Wenjin Tao, Ming-Chuan Leu, Xiaoqing Frank Liu, Rakib Shahriar, S M Nahian Al Sunny
Mechanical and Aerospace Engineering Faculty Research & Creative Works
The common modeling of digital twins uses an information model to describe the physical machines. The integration of digital twins into productive cyber-physical cloud manufacturing (CPCM) systems imposes strong demands such as reducing overhead and saving resources. In this paper, we develop and investigate a new method for building cloud-based digital twins (CBDT), which can be adapted to the CPCM platform. Our method helps reduce computing resources in the information processing center for efficient interactions between human users and physical machines. We introduce a knowledge resource center (KRC) built on a cloud server for information intensive applications. An information model …
Worker Activity Recognition In Smart Manufacturing Using Imu And Semg Signals With Convolutional Neural Networks, Wenjin Tao, Ze-Hao Lai, Ming-Chuan Leu, Zhaozheng Yin
Worker Activity Recognition In Smart Manufacturing Using Imu And Semg Signals With Convolutional Neural Networks, Wenjin Tao, Ze-Hao Lai, Ming-Chuan Leu, Zhaozheng Yin
Mechanical and Aerospace Engineering Faculty Research & Creative Works
In a smart manufacturing system involving workers, recognition of the worker's activity can be used for quantification and evaluation of the worker's performance, as well as to provide onsite instructions with augmented reality. In this paper, we propose a method for activity recognition using Inertial Measurement Unit (IMU) and surface electromyography (sEMG) signals obtained from a Myo armband. The raw 10-channel IMU signals are stacked to form a signal image. This image is transformed into an activity image by applying Discrete Fourier Transformation (DFT) and then fed into a Convolutional Neural Network (CNN) for feature extraction, resulting in a high-level …
High-Frequency Instabilities Of Stationary Crossflow Vortices In A Hypersonic Boundary Layer, Fei Li, Meelan Choudhari, Pedro Paredes-Gonzalez, Lian Duan
High-Frequency Instabilities Of Stationary Crossflow Vortices In A Hypersonic Boundary Layer, Fei Li, Meelan Choudhari, Pedro Paredes-Gonzalez, Lian Duan
Mechanical and Aerospace Engineering Faculty Research & Creative Works
Hypersonic boundary layer flows over a circular cone at moderate incidence angle can support strong crossflow instability in between the windward and leeward rays on the plane of symmetry. Due to more efficient excitation of stationary crossflow vortices by surface roughness, such boundary layer flows may transition to turbulence via rapid amplification of the high-frequency secondary instabilities of finite-amplitude stationary crossflow vortices. The amplification characteristics of these secondary instabilities are investigated for crossflow vortices generated by an azimuthally periodic array of roughness elements over a 7° half-angle circular cone in a Mach 6 free stream. The analysis is based on …
Ionic And Electronic Conductivities Of Atomic Layer Deposition Thin Film Coated Lithium Ion Battery Cathode Particles, Rajankumar L. Patel, Jonghyun Park, Xinhua Liang
Ionic And Electronic Conductivities Of Atomic Layer Deposition Thin Film Coated Lithium Ion Battery Cathode Particles, Rajankumar L. Patel, Jonghyun Park, Xinhua Liang
Mechanical and Aerospace Engineering Faculty Research & Creative Works
It is imperative to ascertain the ionic and electronic components of the total conductivity of an electrochemically active material. A blocking technique, called the “Hebb-Wagner method”, is normally used to explain the two components (ionic and electronic) of a mixed conductor, in combination with the complex ac impedance method and dc polarization measurements. CeO2 atomic layer deposition (ALD)-coated and uncoated, LiMn2O4 (LMO) and LiMn1.5Ni0.5O4 (LMNO) powders were pressed into pellets and then painted with silver to act as a blocking electrode. The electronic conductivities were derived from the currents obtained using …
Geometric Consideration Of Nanostructures For Energy Storage Systems, Jonghyun Park, Jie Li, Wei Lu, Ann Marie Sastry
Geometric Consideration Of Nanostructures For Energy Storage Systems, Jonghyun Park, Jie Li, Wei Lu, Ann Marie Sastry
Mechanical and Aerospace Engineering Faculty Research & Creative Works
Battery performance and its fade are determined by various aspects such as the transport of ions and electrons through heterogeneous internal structures; kinetic reactions at the interfaces; and the corresponding interplay between mechanical, chemical, and thermal responses. The fundamental factor determining this complex multiscale and multiphysical nature of a battery is the geometry of active materials. In this work, we systematically consider the tradeoffs among a selection of limiting geometries of media designed to store ions or other species via a diffusion process. Specifically, we begin the investigation by considering diffusion in spheres, rods, and plates at the particle level, …
Silicon-Wall Interfacial Free Energy Via Thermodynamics Integration, Wan Shou, Heng Pan
Silicon-Wall Interfacial Free Energy Via Thermodynamics Integration, Wan Shou, Heng Pan
Mechanical and Aerospace Engineering Faculty Research & Creative Works
We compute the interfacial free energy of a silicon system in contact with flat and structured walls by molecular dynamics simulation. The thermodynamics integration method, previously applied to Lennard-Jones potentials [R. Benjamin and J. Horbach, J. Chem. Phys. 137, 044707 (2012)], has been extended and implemented in Tersoff potentials with two-body and three-body interactions taken into consideration. The thermodynamic integration scheme includes two steps. In the first step, the bulk Tersoff system is reversibly transformed to a state where it interacts with a structureless flat wall, and in a second step, the flat structureless wall is reversibly transformed into an …
Nonlinear Development And Secondary Instability Of Traveling Crossflow Vortices, Fei Li, Meelan M. Choudhari, Lian Duan, Chau-Lyan Chang
Nonlinear Development And Secondary Instability Of Traveling Crossflow Vortices, Fei Li, Meelan M. Choudhari, Lian Duan, Chau-Lyan Chang
Mechanical and Aerospace Engineering Faculty Research & Creative Works
Building upon the prior research targeting the laminar breakdown mechanisms associated with stationary crossflow instability over a swept-wing configuration, this paper investigates the secondary instability of traveling crossflow modes as an alternate scenario for transition. For the parameter range investigated herein, this alternate scenario is shown to be viable unless the initial amplitudes of the traveling crossflow instability are lower than those of the stationary modes by considerably more than one order of magnitude. The linear growth predictions based on the secondary instability theory are found to agree well with both parabolized stability equations and direct numerical simulation, and the …
Crystallization In Nano-Confinement Seeded By A Nanocrystal -- A Molecular Dynamics Study, Heng Pan, Costas Grigoropoulos
Crystallization In Nano-Confinement Seeded By A Nanocrystal -- A Molecular Dynamics Study, Heng Pan, Costas Grigoropoulos
Mechanical and Aerospace Engineering Faculty Research & Creative Works
Seeded crystallization and solidification in nanoscale confinement volumes have become an important and complex topic. Due to the complexity and limitations in observing nanoscale crystallization, computer simulation can provide valuable details for supporting and interpreting experimental observations. In this article, seeded crystallization from nano-confined liquid, as represented by the crystallization of a suspended gold nano-droplet seeded by a pre-existing gold nanocrystal seed, was investigated using molecular dynamics simulations in canonical (NVT) ensemble. We found that the crystallization temperature depends on nano-confinement volume, crystal orientation, and seed size as explained by classical two-sphere model and Gibbs-Thomson effect.
Object-Oriented Modeling And Fauly Detection Of A Powder Feeder For A Laser Metal Deposition System, Robert G. Landers, Michael Gene Hilgers, Frank W. Liou, Bruce M. Mcmillin
Object-Oriented Modeling And Fauly Detection Of A Powder Feeder For A Laser Metal Deposition System, Robert G. Landers, Michael Gene Hilgers, Frank W. Liou, Bruce M. Mcmillin
Mechanical and Aerospace Engineering Faculty Research & Creative Works
One of the most critical components of a laser metal deposition system is the powder feeder unit. Techniques are required to efficiently design and analyze these systems and to ensure fault tolerance during the operation. In this paper, a model of the powder feeder unit, based on object- oriented abstractions of it components, is developed. This model is modular in that specific components can be efficiently updated or replaced with components that perform a similar function. In addition, the model can be used for simulation allowing for efficient design and analysis. Assurance of the correctness of the powder feeder system …
Object-Oriented Modeling And Fault Detection Of A Powder Feeder For A Laser Metal Deposition System, Robert G. Landers, Michael Gene Hilgers, Frank W. Liou, Bruce M. Mcmillin
Object-Oriented Modeling And Fault Detection Of A Powder Feeder For A Laser Metal Deposition System, Robert G. Landers, Michael Gene Hilgers, Frank W. Liou, Bruce M. Mcmillin
Mechanical and Aerospace Engineering Faculty Research & Creative Works
One of the most critical components of a laser metal deposition system is the powder feeder unit. Techniques are required to efficiently design and analyze these systems and to ensure fault tolerance during the operation. In this paper, a model of the powder feeder unit, based on object-oriented abstractions of it components, is developed. This model is modular in that specific components can be efficiently updated or replaced with components that perform a similar function. In addition, the model can be used for simulation allowing for efficient design and analysis. Assurance of the correctness of the powder feeder system is …
Identification Of Cutting Force In End Milling Operations Using Recurrent Neural Networks, Q. Xu, K. Krishnamurthy, Bruce M. Mcmillin, Wen Feng Lu
Identification Of Cutting Force In End Milling Operations Using Recurrent Neural Networks, Q. Xu, K. Krishnamurthy, Bruce M. Mcmillin, Wen Feng Lu
Mechanical and Aerospace Engineering Faculty Research & Creative Works
The problem of identifying the cutting force in end milling operations is considered in this study. Recurrent neural networks are used here and are trained using a recursive least squares training algorithm. Training results for data obtained from a SAJO 3-axis vertical milling machine for steady slot cuts are presented. The results show that a recurrent neural network can learn the functional relationship between the feed rate and steady-state average resultant cutting force very well. Furthermore, results for the Mackey-Glass time series prediction problem are presented to illustrate the faster learning capability of the neural network scheme presented here
A Recursive Least Squares Training Algorithm For Multilayer Recurrent Neural Networks, Q. Xu, K. Krishnamurthy, Bruce M. Mcmillin, Wen Feng Lu
A Recursive Least Squares Training Algorithm For Multilayer Recurrent Neural Networks, Q. Xu, K. Krishnamurthy, Bruce M. Mcmillin, Wen Feng Lu
Mechanical and Aerospace Engineering Faculty Research & Creative Works
Recurrent neural networks have the potential to perform significantly better than the commonly used feedforward neural networks due to their dynamical nature. However, they have received less attention because training algorithms/architectures have not been well developed. In this study, a recursive least squares algorithm to train recurrent neural networks with an arbitrary number of hidden layers is developed. The training algorithm is developed as an extension of the standard recursive estimation problem. Simulated results obtained for identification of the dynamics of a nonlinear dynamical system show promising results.
Neural Modeling And Control Of A Distillation Column, James Edward Steck, K. Krishnamurthy, Bruce M. Mcmillin, Gary G. Leininger
Neural Modeling And Control Of A Distillation Column, James Edward Steck, K. Krishnamurthy, Bruce M. Mcmillin, Gary G. Leininger
Mechanical and Aerospace Engineering Faculty Research & Creative Works
Control of a nine-stage three-component distillation column is considered. The control objective is achieved using a neural estimator and a neural controller. The neural estimator is trained to represent the chemical process accurately, and the neural controller is trained to give an input to the chemical process which will yield the desired output. Training of both the neural networks is accomplished using a recursive least squares training algorithm implemented on an Intel iPSC/2 multicomputer (hypercube). Simulated results are presented for a numerical example.