System Identification and Response Prediction of Deteriorating Structures

2006 ◽  
Author(s):  
C. H. Ng ◽  
N. Ajavakom ◽  
F. Ma
Keyword(s):  
System Identification ◽  
Real Life ◽  
Response Prediction ◽  
Identification Algorithm ◽  
Hysteretic Model ◽  
Sea Waves ◽  
Major Barrier ◽  
Differential Model ◽  
Successful Design ◽  
Load Displacement

The lack of a fundamental theory of hysteresis is a major barrier to successful design of structures against deterioration associated with earthquakes, high winds, and sea waves. Development of a practical model of degrading structures that would match experimental observations is an important task. This paper has a two-fold objective. First, a superior system identification algorithm is devised to estimate the unspecified parameters in a differential model of hysteresis from experimental load-displacement traces. This algorithm is based upon the latest theory of genetic evolution and it will be streamlined through global sensitivity analysis. Second, the utility of identification of hysteresis is demonstrated through response prediction. Suppose a hysteretic model is generated with a given load-displacement trace. It will be shown experimentally that the model will predict the response of the same system driven by other cyclic loads. The requirements for precise prediction will be addressed. Through identification of hysteresis, it becomes possible to assess, for the first time in analysis, the performance of a real-life structure that has previously been damaged. In the open literature, there is not any other method that can perform the same task.

Identification and Performance Prediction of Nonlinear Degrading Structures

2007 ◽  
Author(s):  
Nopdanai Ajavakom ◽  
Ching H. Ng ◽  
Fai Ma
Keyword(s):  
Real Life ◽  
Response Prediction ◽  
Identification Algorithm ◽  
Hysteretic Model ◽  
Sea Waves ◽  
Major Barrier ◽  
Differential Model ◽  
Successful Design ◽  
Load Displacement ◽  
And Performance

The lack of a fundamental theory of hysteresis is a major barrier to successful design of structures against deterioration associated with earthquakes, high winds, and sea waves. Development of a practical model of degrading structures that would match experimental observations is an important task. This paper has a two-fold objective. First, a superior system identification algorithm is devised to estimate the unspecified parameters in a differential model of hysteresis from experimental load-displacement traces. This algorithm is based upon the latest theory of genetic evolution and it will be streamlined through global sensitivity analysis. Second, the utility of identification of hysteresis is demonstrated through nonlinear response prediction, which is important in structural design. Suppose a hysteretic model is generated with a given load-displacement trace. It will be shown experimentally that the model will predict the response of the same system driven by other cyclic loads. The requirements for accurate prediction will be addressed. Through identification of hysteresis, it becomes possible to assess the performance of a real-life structure that has previously been damaged. In the open literature, there is not any other method that can perform the same task.

On Response Prediction of Degrading Structures

2005 ◽  
Author(s):  
C. H. Ng ◽  
N. Ajavakom ◽  
F. Ma
Keyword(s):  
Structure Prediction ◽  
Hysteresis Loops ◽  
Response Prediction ◽  
Parametric Identification ◽  
Practical Significance ◽  
Identification Algorithm ◽  
Hysteretic Model ◽  
Sea Waves ◽  
Future Performance ◽  
Brute Force Approach

All structures degrade when acted upon by cyclic forces associated with earthquakes, high winds, and sea waves. Identification and prediction of degradation is thus a problem of considerable practical significance in the field of engineering mechanics. Under cyclic excitations, system degradation manifests itself in the evolution of the associated hysteresis loops. In this paper, a robust identification algorithm is devised to generate hysteretic models of a deteriorating structure from its experimental load-displacement traces. This algorithm is based upon the generalized Bouc-Wen model and the latest theory of differential evolution, streamlined through global sensitivity analysis. It can account for strength degradation, stiffness degradation, and pinching characteristics in the evolution of hysteretic traces, whereby earlier studies in parametric identification of hysteresis are extended. In addition, it is shown experimentally that a hysteretic model obtained by identification can be used to predict the future performance of a degrading structure. Prediction of degradation through identification is a brute-force approach that offers a close representation of reality. There is not any method based upon the fundamental postulates of mechanics that can predict the response of a degrading structure well beyond its linear range.

Parameter Identification of Inelastic Structures

1999 ◽  
Author(s):  
H. Zhang ◽  
Y. Yang ◽  
G. C. Foliente ◽  
F. Ma
Keyword(s):  
Parameter Identification ◽  
Gradient Methods ◽  
Model Parameters ◽  
Inelastic Behavior ◽  
Hysteretic Model ◽  
Differential Model ◽  
Inelastic Structures ◽  
Curve Fit ◽  
Successful Design ◽  
Strong Winds

Abstract Structures often exhibit nonlinear and inelastic behavior in the form of hysteresis loop under severe loads associated with earthquake, austere winds and waves. Hysteresis is particularly important in depicting the nonlinear response of wood buildings, braced steel frames, reinforced concrete, and structures with a high proportion of composite materials. A practical model of hysteresis that would match experimental observations on real structures is needed for the successful design of structures against earthquakes and strong winds. Two different time-domain system identification algorithms will be presented in this report to estimate the parameters of an extended Bouc-Wen hysteretic model. This version of the differential model of hysteresis can curve-fit practically any hysteresis trace with a suitable choice of the model parameters. Thirteen control parameters are included in the model. The parameter identification algorithms presented in this report include the constrained simplex and generalized reduced gradient methods. Noise filtering techniques and constraints will also be used in this study to assist in parameter identification. The effectiveness of the proposed algorithms will be demonstrated through simulations of nonlinear systems with pinching and degradation characteristics. Due to very modest computing requirements, the proposed identification algorithms can be acceptable as a basic tool for estimating hysteretic parameters in engineering design.

scholarly journals Nonlinear Aeroelastic System Identification Based on Neural Network

2018 ◽  
Vol 8 (10) ◽  
pp. 1916
Author(s):  
Bo Zhang ◽  
Jinglong Han ◽  
Haiwei Yun ◽  
Xiaomao Chen
Keyword(s):  
Neural Network ◽  
Time Delay ◽  
System Identification ◽  
Control Surface ◽  
Identification Algorithm ◽  
Aeroelastic System ◽  
Structural Nonlinearity ◽  
Time Marching ◽  
Artificial Neural Network Ann ◽  
Trained Neural Network

This paper focuses on the nonlinear aeroelastic system identification method based on an artificial neural network (ANN) that uses time-delay and feedback elements. A typical two-dimensional wing section with control surface is modelled to illustrate the proposed identification algorithm. The response of the system, which applies a sine-chirp input signal on the control surface, is computed by time-marching-integration. A time-delay recurrent neural network (TDRNN) is employed and trained to predict the pitch angle of the system. The chirp and sine excitation signals are used to verify the identified system. Estimation results of the trained neural network are compared with numerical simulation values. Two types of structural nonlinearity are studied, cubic-spring and friction. The results indicate that the TDRNN can approach the nonlinear aeroelastic system exactly.

Research of Fan System Identification Based on Improved Bayes Algorithm

2011 ◽  
Vol 66-68 ◽  
pp. 448-453
Author(s):  
Hai Tao Wang ◽  
Ze Zhang
Keyword(s):  
System Identification ◽  
Dynamic Characteristics ◽  
System Modeling ◽  
Estimation Algorithm ◽  
Parametric Model ◽  
Testing System ◽  
Identification Algorithm ◽  
Environment Protection ◽  
Parameter Estimation Algorithm ◽  
Bayes Algorithm

In every filed of natural science, more and more researchers attach importance to system quantitative analysis, control and prediction. In filed of automatic control, system identification is the extension of system dynamic characteristics testing. System modeling is the basis of system identification, non-parametric model can be obtained by means of dynamic characteristics testing, but parametric model must be established by means of parameter estimation algorithm, which is more prevalent than dynamic characteristics testing. Coal power plant produces more gas and dust, so how to control the fan system plays a very important role in environment protection. We must clarify the parameter of fan system before controlling it. The traditional Bayes identification algorithm is used widely in research and industry, and the effect is relatively good. The paper induces the concept of loss function based on traditional Bayes identification algorithm, and proposes an improved Bayes identification algorithm, which can be applied to fan system identification successfully.

Analysis of a Platform for Measuring Moments and Products of Inertia of Large Vehicles

1976 ◽  
Vol 98 (2) ◽  
pp. 186-195
Author(s):  
D. Orne ◽  
T. Schmitz
Keyword(s):  
System Identification ◽  
Test Data ◽  
Normal Modes ◽  
Free Vibration Analysis ◽  
Identification Algorithm ◽  
Test Configuration ◽  
Truncation Errors ◽  
Sample Test ◽  
Instantaneous Center ◽  
Cable Lines

A rigid platform symmetrically supported by four sloping cables is proposed for measuring the center-of-gravity coordinates and the moments and products of inertia of large vehicles such as buses, trucks, and trailers. In addition to a torsional degree-of-freedom, the system undergoes pitch and roll motions about axes through the system instantaneous center which lies directly below the center of the platform at the intersection of the cable lines-of-action under quiescent conditions. The natural frequencies and normal modes of the freely vibrating loaded platform are used as inputs to a linearized System Identification Algorithm for computing the inertia properties of the test vehicle. Hypothetical test data generated from the Free Vibration Analysis of a sample test configuration are used to evaluate the sensitivity of the System Identification Algorithm to inaccuracies in test data or to truncation errors in computation.

Sign in / Sign up

Export Citation Format

Share Document