CFD-based reduced-order modeling of fluidized-bed biomass fast pyrolysis using artificial neural network

2020 ◽  
Vol 152 ◽  
pp. 613-626 ◽  
Author(s):  
Hanbin Zhong ◽  
Qingang Xiong ◽  
Lina Yin ◽  
Juntao Zhang ◽  
Yuqin Zhu ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Fluidized Bed ◽  
Fast Pyrolysis ◽  
Reduced Order Modeling ◽  
Reduced Order ◽  
Artificial Neural ◽  
Biomass Fast Pyrolysis

Reduced-Order Model with an Artificial Neural Network for Aerostructural Design Optimization

2013 ◽  
Vol 50 (4) ◽  
pp. 1106-1116 ◽  
Author(s):  
Kyung Hyun Park ◽  
Sang Ook Jun ◽  
Sung Min Baek ◽  
Maeng Hyo Cho ◽  
Kwan Jung Yee ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Design Optimization ◽  
Reduced Order Model ◽  
Order Model ◽  
Reduced Order ◽  
Artificial Neural

LIMIT CYCLE OSCILLATION PREDICTION AND CONTROL DESIGN METHOD FOR AEROELASTIC SYSTEM BASED ON NEW NONLINEAR REDUCED ORDER MODEL

2011 ◽  
Vol 08 (01) ◽  
pp. 77-90 ◽  
Author(s):  
GANG CHEN ◽  
YUE-MING LI ◽  
GUI-RONG YAN
Keyword(s):  
Limit Cycle ◽  
Active Control ◽  
Reduced Order Model ◽  
Nonlinear Flow ◽  
Limit Cycle Oscillation ◽  
Control Law ◽  
Order Model ◽  
Reduced Order ◽  
Large Perturbation ◽  
Cycle Oscillation

When the amplitude of the oscillation of the unsteady flow is large or there is large perturbation relative to the mean background flow, the traditional proper orthogonal decomposition/reduced order model (POD/ROM) based on linearized time or frequency domain small disturbance solvers cannot capture the main nonlinear features well such as limit cycle oscillation (LCO), which is very dangerous for the structure. Therefore, the traditional linear ROMs are not good enough for limit cycles prediction and active control law design. A new nonlinear ROM based on dynamically nonlinear flow equation NPOD/ROM was investigated. The nonlinear second-order snapshot equation in time domain for POD basis construction is obtained from the Taylor series expansion of the flow solver. The simulation results indicate that the NPOD/ROM can capture LCO very well and is also very convenient for active control law design, while the traditional POD/ROM lose effectiveness.

Reduced Order Modeling of Limit Cycle Oscillation for Aeroelasticity Systems

2002 ◽  
Author(s):  
Philip S. Beran ◽  
David J. Lucia ◽  
Chris L. Pettit
Keyword(s):  
Limit Cycle ◽  
Degrees Of Freedom ◽  
Galerkin Approximation ◽  
Computational Cost ◽  
Coupled System ◽  
Low Rank ◽  
Limit Cycle Oscillation ◽  
Aeroelastic System ◽  
Reduced Order ◽  
Cycle Oscillation

Limit-cycle oscillations of a nonlinear panel in supersonic flow are computed using a reduced-order aeroelastic model. Panel dynamics are governed by the large-deflection, nonlinear, von Ka´rma´n equation as expressed in low-order form through a Galerkin approximation. The aerodynamics are described by the Euler equations, which are reduced in order using proper orthogonal decomposition. The coupled system of equations is implicitly time integrated with second-order temporal accuracy to predict limit-cycle oscillation (LCO) amplitude, and linearly analyzed to predict LCO onset. The fluid is synchronized with the structure in time through subiteration, using only 18 degrees of freedom to describe the aeroelastic system. The Jacobian employed in the fully implicit analysis is of equivalently low rank, enabling rapid analysis. Using the reduced order model, LCO onset is predicted directly at a computational cost of approximately 400 time steps with a high accuracy verified by full-order analysis.

Sign in / Sign up

Export Citation Format

Share Document