Keywords: Chao theory, dynamic neural network, wavelet, fuzzy logic, genetic algorithm, structural health monitoring, nonlinear active control, highrise building.

In this chapter examples of recent and current research in the areas of vibration control and health monitoring of smart structures performed by the author and his research associates are presented. In such structures sensors monitor the health of the structure and properly-designed actuators apply internal forces to compensate for the forces of nature due to earthquakes and winds. Multidisciplinary methodologies and innovative computational models are presented for health monitoring and nonlinear active control of civil structures subject to extreme dynamic loadings. The methodologies are based on adroit integration of multiple computing paradigms including chaos theory, wavelets, and three soft computing methods, fuzzy logic, neural networks, and genetic algorithm.

A multi-paradigm dynamic time-delay fuzzy wavelet neural network (WNN) model is presented for nonparametric identification of structures with nonlinear behaviour using the nonlinear autoregressive moving average with exogenous inputs. Noise in the signals is removed using the discrete wavelet packet transform method. In order to preserve the dynamics of time series, the reconstructed state space concept from the chaos theory is employed to construct the input vector. In addition to denoising, wavelets are employed in combination with neural networks and fuzzy logic to create a new pattern recognition model to capture the characteristics of the time series sensor data accurately and efficiently.

Compared with conventional neural networks, training of a dynamic neural network for system identification of large-scale structures is substantially more complicated and time-consuming because both input and output of the network are not single-valued but thousands of time steps. An adaptive Levenberg-Marquardt-least squares (LM-LS) algorithm with a backtracking inexact linear search scheme is presented for training of the dynamic fuzzy WNN model.

A nonparametric system identification-based model is developed for damage detection of irregular highrise building structures using the dynamic fuzzy WNN model with an adaptive LM-LS learning algorithm. A power density spectrum method is proposed for damage detection in a structure. The damage detection methodology is validated using the sensed data obtained for a 38-story concrete test model. This research provides an effective tool for real-time health monitoring and nondestructive damage evaluation of both highrise building and bridge structures.

A new nonlinear control model is presented for active control of three dimensional highrise building structures under extreme dynamic loadings. Both material and geometrical nonlinearities are considered in modelling the structural response. A dynamic fuzzy WNN is developed as a fuzzy wavelet neuroemulator to predict structural responses from the immediate past structural responses and actuator dynamics. A floating-point genetic algorithm is developed for finding the optimal control forces for active nonlinear control of building structures using the dynamic fuzzy wavelet neuroemulator. The algorithm does not need the pre-training required in a neural network-based controller, which improves the efficiency of the general control methodology significantly. A highrise steel building structure with vertical irregularity is used to validate the new neuro-genetic control algorithm under two seismic excitations. Validation results demonstrate that the new control methodology is effective in significantly reducing the response of large irregular building structures subjected to seismic excitations.

purchase the full-text of this chapter (price £20)

go to the previous chapter
go to the next chapter
return to the table of contents
return to the book description
purchase this book (price £98 +P&P)