Analysis and modification of Volterra/Wiener neural networks for the adaptive identification of non-linear hysteretic dynamic systems

Pei, Jin‐Song; Smyth, Andrew W.; Kosmatopoulos, Elias B.

doi:10.1016/j.jsv.2003.06.005

Cited by 52 publications

(25 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…al. [12], this study will focus on the use of the VWNN to predict accelerations for a system with defined degrees-of-freedom excited by a forcing function.…”

Section: Volterra-weiner Neural Network (Vwnn)mentioning

confidence: 99%

See 1 more Smart Citation

Distributed neural computations for embedded sensor networks

2011

View full text Add to dashboard Cite

Wireless sensing technologies have recently emerged as an inexpensive and robust method of data collection in a variety of structural monitoring applications. In comparison with cabled monitoring systems, wireless systems offer low-cost and low-power communication between a network of sensing devices. Wireless sensing networks possess embedded data processing capabilities which allow for data processing directly at the sensor, thereby eliminating the need for the transmission of raw data. In this study, the Volterra/Weiner neural network (VWNN), a powerful modeling tool for nonlinear hysteretic behavior, is decentralized for embedment in a network of wireless sensors so as to take advantage of each sensor's processing capabilities. The VWNN was chosen for modeling nonlinear dynamic systems because its architecture is computationally efficient and allows computational tasks to be decomposed for parallel execution. In the algorithm, each sensor collects it own data and performs a series of calculations. It then shares its resulting calculations with every other sensor in the network, while the other sensors are simultaneously exchanging their information. Because resource conservation is important in embedded sensor design, the data is pruned wherever possible to eliminate excessive communication between sensors. Once a sensor has its required data, it continues its calculations and computes a prediction of the system acceleration. The VWNN is embedded in the computational core of the Narada wireless sensor node for on-line execution. Data generated by a steel framed structure excited by seismic ground motions is used for validation of the embedded VWNN model.

show abstract

“…al. [12], this study will focus on the use of the VWNN to predict accelerations for a system with defined degrees-of-freedom excited by a forcing function.…”

Section: Volterra-weiner Neural Network (Vwnn)mentioning

confidence: 99%

“…al. [12], has a unique architecture that allows for simple decomposition onto a WSN. There are several benefits for embedding the neural network across a network of wireless sensors.…”

Section: Introductionmentioning

confidence: 99%

Distributed neural computations for embedded sensor networks

2011

View full text Add to dashboard Cite

show abstract

“…The literature provides an extensive coverage of the techniques used for identifying non-linear and hysteretic systems, based on different types of excitation and numerical algorithms: in general, methods can be classified as belonging to either the parametric approach (see for example [23,24]) or the non parametric approach (see for example [25][26][27]). In the former case, a priori selection of a specific model for the dynamic behaviour of the system is needed and the identification process consists of determining the coefficients for such model; non parametric methods, instead, do not require any assumption as to the type and localisation of structural non-linearities, but the identified quantities cannot be generally correlated to the system's equation of motion (for instance, neural network based methods).…”

Section: Polynomial Time-varying Identificationmentioning

confidence: 99%

Non-linear Damping and Frequency Identification in a Progressively Damaged R.C. Element

Demarie

Sabia

2010

Exp Mech

View full text Add to dashboard Cite

This paper describes the non-linear identification of a progressively damaged reinforced concrete beam-column node. The aims are the detection and identification of the different sources of damping and their dependence from the damage level. To this end a specially formulated non-linear identification method is proposed, based on a time-varying polynomial approximation of the system dynamics, suitable for use in the presence of excitations of any form. A minimum condition imposed to the identified dissipated energy leads to the distinction of the linear viscous component from the other damping mechanisms. The estimated values obtained from the experimental tests show a significant influence of the damage level on the linear viscous damping coefficient. This suggests that, in a non-linear dynamic time-history analysis, the use of Rayleigh damping model with proportionality to the initial stiffness is basically in contrast with experimental evidence and more refined viscous damping models are needed for prediction purposes.

show abstract

“…The proposed model is composed of hysteresis-like subsystem and dynamic non-linear subsystem. Based on models presented by Serpico [8] and Pei [12], the proposed model is written as yðt þ 1Þ ¼ Q½yðtÞ; P u ðuðtÞ; yðtÞÞ; uðtÞ 8uX0,…”

Section: Neural Network Hysteresis Model Based On Hybrid Modelmentioning

confidence: 99%