A combined neural network/gradient-based approach for the identification of constitutive model parameters using self-boring pressuremeter tests

2009 ◽  
Vol 33 (6) ◽  
pp. 817-849 ◽  
Author(s):  
Rafał F. Obrzud ◽  
Laurent Vulliet ◽  
Andrzej Truty
Keyword(s):  
Neural Network ◽  
Constitutive Model ◽  
Model Parameters ◽  
Gradient Based

Intelligent Identification of High Rock-Filled Embankment Constitutive Model Parameters

2011 ◽  
Vol 183-185 ◽  
pp. 2139-2142 ◽  
Author(s):  
Cheng Zhong Yang ◽  
Xian Da Xu
Keyword(s):  
Neural Network ◽  
Constitutive Model ◽  
Design Method ◽  
Orthogonal Design ◽  
Model Parameters ◽  
Maximum Relative Error ◽  
The Finite Element Method ◽  
Orthogonal Design Method ◽  
Design Requirements ◽  
Inverse Analysis Method

Based on the orthogonal design method, the finite element method was combined with the artificial neural network to have established high rock-filled embankment constitutive model parameters inverse analysis method. According to orthogonal design requirements, the level of inversion parameters were selected and the numerical simulation program were determined. By ANSYS software calculated out the analysis samples of neural network and trained the BP neural network.Using the field monitoring displacement,the soil constitutive model parameters were identified and the inversion parameters were compared with the theoretical value.The results show that: the maximum relative error of the inversion value with the theoretical value is no more than 9%,which meets accuracy requirements.

Parameter Identification of Soil Hyperbolic Constitutive Model by Inverse Analysis Procedure

2007 ◽  
Vol 340-341 ◽  
pp. 1231-1236
Author(s):  
Shou Ju Li ◽  
Ying Xi Liu ◽  
Hai Yun Cao ◽  
Dong Cheng
Keyword(s):  
Neural Network ◽  
Constitutive Model ◽  
Parameter Identification ◽  
Inverse Analysis ◽  
Calculated Data ◽  
Soil Mass ◽  
Analysis Procedure ◽  
Identification Accuracy ◽  
Model Parameters ◽  
Construction Simulation

A tangent modulus of soil mass which allows for a piece-wise linear approximation of the hyperbolic response curve is particularly suited for incremental construction simulation. The parameter identification of nonlinear constitutive model of soil mass is based on an inverse analysis procedure, which consists of minimizing the objective function representing the difference between the experimental data and the calculated data of the mechanical model. The artificial neural network is applied to estimate the model parameters of soil mass. The weights of neural network are trained by using the Levenberg-Marquardt approximation which has a fast convergent ability. The parameter identification results illustrate that the proposed neural network has not only higher computing efficiency but also better identification accuracy. The numerically computational results with finite element method show that the forecasted displacements at observing points according to identified model parameters can precisely agree with the observed displacements.

scholarly journals On the optimisation efficiency for the inverse identification of constitutive model parameters

2021 ◽  
Author(s):  
Miguel Guimarães Oliveira ◽  
João Miguel Peixoto Martins ◽  
Bernardete Coelho ◽  
Sandrine Thuillier ◽  
António Andrade-Campos
Keyword(s):  
Constitutive Model ◽  
Mechanical Tests ◽  
Model Parameters ◽  
Inverse Identification ◽  
Full Field ◽  
Material Parameters ◽  
Gradient Based ◽  
Identification Techniques ◽  
Optimisation Algorithms ◽  
Optimisation Technique

The development of full-field measurement techniques paved the way for the design of new mechanical tests. However, because these mechanical tests provide heterogeneous strain fields, no closed-form solution exists between the measured deformation fields and the constitutive parameters. Therefore, inverse identification techniques should be used to calibrate constitutive models, such as the widely known finite element model updating (FEMU) and the virtual fields method (VFM). Although these inverse identification techniques follow distinct approaches to explore full-field measurements, they all require using an optimisation technique to find the optimum set of material parameters. Nonetheless, the choice of a suitable optimisation technique lacks attention and proper research. Most studies tend to use a least-squares gradient-based optimisation technique, such as the Levenberg-Marquardt algorithm. This work analyses optimisation algorithms, gradient-based and -free algorithms, for the inverse identification of constitutive model parameters. To avoid needless implementation and take advantage of highly developed programming languages, the optimisation algorithms available in optimisation libraries are used. A FEMU based approach is considered in the calibration of a thermoelastoviscoplastic model. The material parameters governing strain hardening, temperature and strain rate are identified. Results are discussed in terms of efficiency and the robustness of the optimisation processes.

scholarly journals A General Temperature-Dependent Stress–Strain Constitutive Model for Polymer-Bonded Composite Materials

Polymers ◽  
2021 ◽  
Vol 13 (9) ◽  
pp. 1393
Author(s):  
Xiaochang Duan ◽  
Hongwei Yuan ◽  
Wei Tang ◽  
Jingjing He ◽  
Xuefei Guan
Keyword(s):  
Composite Materials ◽  
Constitutive Model ◽  
Model Parameters ◽  
Temperature Dependent ◽  
Stress Strain ◽  
Proposed Model ◽  
Single Temperature ◽  
Testing Data ◽  
Bonded Composite ◽  
Tension And Compression

This study develops a general temperature-dependent stress–strain constitutive model for polymer-bonded composite materials, allowing for the prediction of deformation behaviors under tension and compression in the testing temperature range. Laboratory testing of the material specimens in uniaxial tension and compression at multiple temperatures ranging from −40 ∘C to 75 ∘C is performed. The testing data reveal that the stress–strain response can be divided into two general regimes, namely, a short elastic part followed by the plastic part; therefore, the Ramberg–Osgood relationship is proposed to build the stress–strain constitutive model at a single temperature. By correlating the model parameters with the corresponding temperature using a response surface, a general temperature-dependent stress–strain constitutive model is established. The effectiveness and accuracy of the proposed model are validated using several independent sets of testing data and third-party data. The performance of the proposed model is compared with an existing reference model. The validation and comparison results show that the proposed model has a lower number of parameters and yields smaller relative errors. The proposed constitutive model is further implemented as a user material routine in a finite element package. A simple structural example using the developed user material is presented and its accuracy is verified.

Gradient-based multi-focus image fusion method using convolution neural network

2021 ◽  
Vol 92 ◽  
pp. 107174
Author(s):  
Yang Zhou ◽  
Xiaomin Yang ◽  
Rongzhu Zhang ◽  
Kai Liu ◽  
Marco Anisetti ◽  
...  
Keyword(s):  
Neural Network ◽  
Image Fusion ◽  
Convolution Neural Network ◽  
Fusion Method ◽  
Gradient Based ◽  
Focus Image ◽  
Image Fusion Method

scholarly journals Enhanced Neural Network Model for Worldwide Estimation of Weighted Mean Temperature

2021 ◽  
Vol 13 (12) ◽  
pp. 2405
Author(s):  
Fengyang Long ◽  
Chengfa Gao ◽  
Yuxiang Yan ◽  
Jinling Wang
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Model Performance ◽  
Global Scale ◽  
Radiosonde Data ◽  
Model Parameters ◽  
Measured Temperature ◽  
Weighted Mean ◽  
Weighted Mean Temperature ◽  
Application Scope

Precise modeling of weighted mean temperature (Tm) is critical for realizing real-time conversion from zenith wet delay (ZWD) to precipitation water vapor (PWV) in Global Navigation Satellite System (GNSS) meteorology applications. The empirical Tm models developed by neural network techniques have been proved to have better performances on the global scale; they also have fewer model parameters and are thus easy to operate. This paper aims to further deepen the research of Tm modeling with the neural network, and expand the application scope of Tm models and provide global users with more solutions for the real-time acquisition of Tm. An enhanced neural network Tm model (ENNTm) has been developed with the radiosonde data distributed globally. Compared with other empirical models, the ENNTm has some advanced features in both model design and model performance, Firstly, the data for modeling cover the whole troposphere rather than just near the Earth’s surface; secondly, the ensemble learning was employed to weaken the impact of sample disturbance on model performance and elaborate data preprocessing, including up-sampling and down-sampling, which was adopted to achieve better model performance on the global scale; furthermore, the ENNTm was designed to meet the requirements of three different application conditions by providing three sets of model parameters, i.e., Tm estimating without measured meteorological elements, Tm estimating with only measured temperature and Tm estimating with both measured temperature and water vapor pressure. The validation work is carried out by using the radiosonde data of global distribution, and results show that the ENNTm has better performance compared with other competing models from different perspectives under the same application conditions, the proposed model expanded the application scope of Tm estimation and provided the global users with more choices in the applications of real-time GNSS-PWV retrival.

scholarly journals A Flow Stress Model of 300M Steel for Isothermal Tension

Materials ◽  
2021 ◽  
Vol 14 (2) ◽  
pp. 252
Author(s):  
Rongchuang Chen ◽  
Shiyang Zhang ◽  
Xianlong Liu ◽  
Fei Feng
Keyword(s):  
Flow Stress ◽  
Constitutive Model ◽  
Tensile Deformation ◽  
Flow Behavior ◽  
Tensile Tests ◽  
Uniaxial Tensile ◽  
Model Parameters ◽  
Average Deviation ◽  
Cross Sectional ◽  
300M Steel

To investigate the effect of hot working parameters on the flow behavior of 300M steel under tension, hot uniaxial tensile tests were implemented under different temperatures (950 °C, 1000 °C, 1050 °C, 1100 °C, 1150 °C) and strain rates (0.01 s−1, 0.1 s−1, 1 s−1, 10 s−1). Compared with uniaxial compression, the tensile flow stress was 29.1% higher because dynamic recrystallization softening was less sufficient in the tensile stress state. The ultimate elongation of 300M steel increased with the decrease of temperature and the increase of strain rate. To eliminate the influence of sample necking on stress-strain relationship, both the stress and the strain were calibrated using the cross-sectional area of the neck zone. A constitutive model for tensile deformation was established based on the modified Arrhenius model, in which the model parameters (n, α, Q, ln(A)) were described as a function of strain. The average deviation was 6.81 MPa (6.23%), showing good accuracy of the constitutive model.

Study on Multi-Axial Mechanical Properties of a Polyurethane Foam and Experimental Verification

2011 ◽  
Vol 311-313 ◽  
pp. 301-308
Author(s):  
Shou Hong Han ◽  
Zhen Hua Lu ◽  
Yong Jin Liu
Keyword(s):  
Mechanical Properties ◽  
Constitutive Model ◽  
Polyurethane Foam ◽  
Axial Compression ◽  
Axial Loading ◽  
Hydrostatic Compression ◽  
Model Parameters ◽  
Loading Conditions ◽  
Three Point Bending ◽  
Pu Foam

In order to investigate the multi-axial mechanical properties of a kind of PU (polyurethane) foam, some experiments in different loading conditions including uni-axial tension, uni-axial compression, hydrostatic compression and three-point bending were conducted. It is shown that the hydrostatic component influences yield behavior of PU foam, the yield strength and degree of strain hardening in hydrostatic compression exceed those for uni-axial compression. In terms of the differential hardening constitutive model, the evolution of PU foam yield surface and plastic hardening laws were fitted from experimental data. A finite element method was applied to analyze the quasi-static responses of the PU foam sandwich beam subjected to three-point bending, and good agreement was observed between experimental load-displacement responses and computational predictions, which validated the multi-axial loading methods and stress-strain constitutive model parameters. Moreover, effects of two foam models applied to uni-axial loading and multi-axial loading conditions were analyzed and compared with three-point bending tests and simulations. It is found that the multi-axial constitutive model can bring more accurate prediction whose parameters are obtained from the tests above mentioned.

scholarly journals Fpga Implementation of Precise Convolutional Neural Network for Extreme Learning Machine

2020 ◽  
Vol 9 (8) ◽  
pp. 470-480
Keyword(s):  
Neural Network ◽  
Receptive Field ◽  
Convolutional Neural Network ◽  
Computation Time ◽  
Field Approach ◽  
Pixel Array ◽  
Gradient Based ◽  
Feed Forward Neural Networks ◽  
Learning Machine ◽  
Hidden Layer

Feed-forward neural networks can be trained based on a gradient-descent based backpropagation algorithm. But, these algorithms require more computation time. Extreme Learning Machines (ELM’s) are time-efficient, and they are less complicated than the conventional gradient-based algorithm. In previous years, an SRAM based convolutional neural network using a receptive – field Approach was proposed. This neural network was used as an encoder for the ELM algorithm and was implemented on FPGA. But, this neural network used an inaccurate 3-stage pipelined parallel adder. Hence, this neural network generates imprecise stimuli to the hidden layer neurons. This paper presents an implementation of precise convolutional neural network for encoding in the ELM algorithm based on the receptive - field approach at the hardware level. In the third stage of the pipelined parallel adder, instead of approximating the output by using one 2-input 15-bit adder, one 4-input 14-bit adder is used. Also, an additional weighted pixel array block is used. This weighted pixel array improves the accuracy of generating 128 weighted pixels. This neural network was simulated using ModelSim-Altera 10.1d and synthesized using Quartus II 13.0 sp1. This neural network is implemented on Cyclone V FPGA and used for pattern recognition applications. Although this design consumes slightly more hardware resources, this design is more accurate compared to previously existing encoders

Sign in / Sign up

Export Citation Format

Share Document