Dynamic Fuzzy Wavelet Neuroemulator For Non-Linear Control of Irregular Buildins Structures

INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING
Int. J. Numer. Meth. Engng 2008; 74:10451066

Published online 6 December 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/nme.2195
Dynamic fuzzy wavelet neuroemulator for non-linear control of

irregular building structures
Xiaomo Jiang1 and Hojjat Adeli2, ,
1 Department
of Civil and Environmental Engineering, Vanderbilt University, Nashville, TN 37235, U.S.A.

of Civil and Environmental Engineering and Geodetic Science, The Ohio State University,
470 Hitchcock Hall, 2070 Neil Ave, Columbus, OH 43210, U.S.A.
2 Department
SUMMARY
A new non-linear control model is presented for active control of three-dimensional (3D) building
structures. Both geometrical and material non-linearities are included in the structural control formulation.
A dynamic fuzzy wavelet neuroemulator is presented for predicting the structural response in future time
steps. Two dynamic coupling actions are taken into account simultaneously in the control model: (a)
coupling between lateral and torsional motions of the structure and (b) coupling between the actuator and
the structure. The new neuroemulator is validated using two irregular 3D steel building structures, a 12story structure with vertical setbacks and an 8-story structure with plan irregularity. Numerical validations
in both time and frequency domains demonstrate that the new neuroemulator provides accurate prediction
of structural displacement responses, which is required in neural network models for active control of
structures. In the companion paper, a floating-point genetic algorithm is presented for finding the optimal
control forces needed for active non-linear control of building structures using the dynamic fuzzy wavelet
neuroemulator presented in this paper. Copyright q 2007 John Wiley & Sons, Ltd.
Received 5 November 2006; Revised 27 July 2007; Accepted 22 August 2007
KEY WORDS:
non-linear control; wavelet neural network; neuroemulator; active control
INTRODUCTION
Over the past two decades, a large amount of research has been conducted on the development and
implementation of active, semi-active, and hybrid control of structures (e.g. [123]), and various
control strategies have been proposed, such as sliding mode control [22] and optimal polynomial
control [23]. However, most of the studies have focused on the application of classical linear
control theories, such as linear quadratic regulator (LQR) feedback control algorithm [6] and linear
Correspondence
to: Hojjat Adeli, Department of Civil and Environmental Engineering and Geodetic Science,
The Ohio State University, 470 Hitchcock Hall, 2070 Neil Ave, Columbus, OH 43210, U.S.A.
E-mail: adeli.1@osu.edu
Copyright q
2007 John Wiley & Sons, Ltd.
1046
X. JIANG AND H. ADELI
quadratic Gaussian (LQG) control algorithm [2]. These algorithms control structural responses
effectively only when the structure is small and assumed to behave linearly.
There are several motivations for considering active non-linear control of structures. The linear
control algorithms are not effective for active non-linear control of highrise building structures
subjected to extreme loadings. Highrise building structures may experience yielding and non-linear
behavior (geometrical or material non-linearity or both) under severe earthquakes or strong winds.
Any structural damage changes the structural stiffnesses during the extreme dynamic event. In
such cases, a computational model based on the assumption that the controlled structure behaves
linearly would be inadequate for representing the actual behavior of the structure. Linear control
algorithms cannot be used to control the response of structures in the non-linear range effectively
[17]. On the other hand, maintaining a linear behavior for a large controlled structure such as
a highrise building during an extreme dynamic event would require actuators with impractically
large capacities.
Neural networks are known for their ability to model complex and non-linear phenomena where
no explicit mathematical model exists [2426] and for their adaptability for handling incomplete
information. In the past decade or so, a number of articles have been published on the development
and application of neural network-based or other adaptive/intelligent control algorithms for active
linear/non-linear control of mostly small structural systems [8, 1618, 2732]. Ghaboussi and
Joghataie [28] first use a neural network-based emulator to identify the response of a 3-story
two-dimensional (2D) frame structure. Then, a neural network-based controller is trained using
the emulator for linear control of the structure. Structural displacement and acceleration responses
during the previous two time steps and actuator electric signals during the previous three time steps
are used as inputs to the neural network model. Bani-Hani and Ghaboussi [16, 17] extended the
neural network-based algorithm for non-linear control of a benchmark 3-story 2D frame structure
considering material non-linearity. The authors show that the neural network-based approaches
provide adequate accuracy for active control of such small 2D frame structures.
Recently, a few journal articles have been published on active control of three-dimensional
(3D) highrise building structures subjected to earthquake loads (e.g. [3337]). Fur et al. [33] use
an active tuned mass damper (ATMD) system for an 8-story building structure considering the
lateraltorsional coupling. The control gain, an important parameter required in feedback control
algorithms such as LQR, is obtained through the complete feedback of position and velocity.
Al-Dawod et al. [34] present a fuzzy logic-based controller for active control of 3- and 20-story
3D symmetric building structures. Ohtori et al. [35] define 3-, 9-, and 20-story symmetric steel
structures for benchmark active control studies without presenting any active non-linear control
algorithm. The 9-story structure has also been investigated in a semi-active control study using
magnetorheological dampers [36] and an active control study using H2 /LQG controller [37].
Recently, Adeli and Kim [20] presented a novel wavelet-hybrid feedback-linear mean square
(LMS) algorithm for robust control of civil structures. It is shown that the wavelet transform can be
used to enhance the performance of feedback control algorithms. An active mass driver benchmark
structure is used to validate the effectiveness of the proposed control model. Simulation results
demonstrate that the proposed model is effective for control of both steady and transient vibrations
without any significant additional computational burden. Kim and Adeli [38] present a hybrid
control system through judicious combination of a passive supplementary damping system with
a semi-active tuned liquid column damper (TLCD) system. The wavelet-hybrid feedback LMS
control algorithm [20] is used to find optimal values of the control parameters for an 8-story frame
using three different simulated earthquake ground accelerations. The wavelet-hybrid feedback LMS
Copyright q

DOI: 10.1002/nme
WAVELET NEUROEMULATOR FOR NON-LINEAR CONTROL OF IRREGULAR STRUCTURES
1047
algorithm has also been used for vibration control of cable-stayed bridges under various seismic
excitations [39].
Recent literature indicates that adaptive/intelligent control algorithms, such as those developed
by Adeli and Kim [20], are advantageous over classical feedback control algorithms for active or
hybrid control of structures for several reasons. First, these algorithms can tolerate the imprecision
in the sensed data. Second, they require less prior knowledge of the structural system to be
controlled. Third, they can be used to handle non-linearity. Finally, they usually converge quickly
and are therefore practical for online active control of large-scale structures.
The focus of this study is active non-linear control of 3D building structures based on the
wavelet neural network (WNN) model developed by authors. Both material and geometrical nonlinearities are considered in modeling the structural response under strong earthquake loadings.
Furthermore, the structural modeling takes into account two coupling actions: the coupling action
between the actuators and the structure and the coupling between the lateral and torsional motions
of 3D irregular structures. A dynamic fuzzy WNN is developed as a fuzzy wavelet neuroemulator
to predict structural responses from the immediate past structural responses and actuator dynamics.
In the companion paper, a floating-point genetic algorithm is presented for finding the optimal
control forces for active non-linear control of building structures using the dynamic fuzzy wavelet
neuroemulator presented in this paper [40].
COUPLED NON-LINEAR DYNAMICS OF 3D BUILDING STRUCTURES
Non-linear dynamics of irregular 3D buildings with an active control system
A 3D highrise building structure can in general have both plan and elevation irregularities. In
general, the center of mass, CM , does not coincide with the center of resistance, CR , in each
floor (Figure 1(a)), and the centers of mass and resistance of floors do not lie on the same
vertical axes (their locations can vary from floor to floor) due to different story stiffnesses along
the two principal directions. This difference may exist even for a structure with a symmetric
plan because different members with different stiffnesses may be used in a symmetric floor
arrangement. In such situations there is a coupling between lateral and torsional motions. The
coupling may result in the maximum lateral displacement in a direction other than the direction of minimum stiffness. In this case, neglecting the coupling between lateral and torsional
motions usually underestimates the maximum responses of the structure under earthquake
loadings [41].
In this study, floor diaphragms are assumed to be rigid and axial deformations of the columns are
neglected. As such, the irregular building structure is modeled using three displacement degrees
of freedom (DOFs) for each floor: translations in x- and y-directions and a rotation about the
vertical axis z passing through the center of resistance. The total number of DOFs is therefore
m = 3L, where L is the number of stories. The structural displacement vector at time t is expressed
as
u(t) = [u 1 (t) v1 (t) 1 (t) u 2 (t) v2 (t) 2 (t) u L (t) vL (t) L (t)]T
(1)
where u i (t), vi (t), and i (t) are the translations in x- and y-directions and the rotation about the
vertical axis z of the ith floor, respectively. The superscript T represents the transpose of the
matrix. The equation of motions of 3D building structures with an active control system subjected
Copyright q

DOI: 10.1002/nme
1048
z
Center of
Mass
Center of
Resistance
y y
CR
CM
x
y
y
x
CR
CM
(b)
Actuator
(a)
Seismic excitation xg(t)
Figure 1. Structural model of a 3D building: (a) 3D building and (b) typical story with four actuators.
to seismic excitations is expressed as
Mu(t)+C
u(t)+R(x,
t) = Ic F(t)M0 Ig x g (t)
(2)
where M and C represent m m mass and damping coefficient matrices of the structure, respectively; R(x, t) is the m 1 restoring force vector; F(t) is the r 1 control force vector whose
elements F(t) are in the form of time series; Ic is an m r location matrix representing the location
of the actuators; and x g (t) is the input horizontal earthquake acceleration which can have any
arbitrary orientation. In this study, we assume that two pairs of actuators are used in every floor
of the building along two perpendicular axes x and y (Figure 1(b)). Each pair consists of two
identical actuators, representing the control force F(t) in the corresponding direction. In practical
applications, the moment caused by actuators in two directions can be minimized by properly
assigning the control force to the two actuators in each direction. For the sake of simplifying
the computations, we assume that the distance between the center of mass (CM ) and the center
of resistance (CR , ) in each floor is very small. Thus, the moment caused by actuators in two
directions is ignored. The focus of the current paper is on novel computational techniques, not
actuator placement. Thus, the proposed methodologies are not tied to any particular configuration.
As such, r = 2L. The angle of orientation of the actuators is = 0 for one pair and = 90 for
another pair. The total number of actuators used in the building is therefore 4L.
Copyright q

DOI: 10.1002/nme
1049
In Equation (2), M0 is an m m diagonal mass matrix whose diagonal terms are the same as the
diagonal terms of the full matrix M, and Ig is an m 1 orientation matrix denoting the orientation
of the external earthquake excitation:
Ig = [cos sin 0 cos sin 0 cos sin 0]T
(3)
where is the direction angle of the earthquake motion measured from the x-axis (Figure 1). The
location matrix Ic in Equation (2) is expressed as follows:
cos 3 sin 3
cos r 1 sin r 1
cos 1 sin 1
cos r
Ic = diag sin 2 cos 2 sin 4 cos 4 sin r
(4)
0
0
0
0
0
0
mr
Figure 2 shows a building structure with an active control system. The earthquake excitation is
represented by the horizontal ground acceleration in the form of a time series, x g (t). The active
control system consists of a computer, hydraulic actuators, and sensors. The computer is used to
receive, process, and send signals between the actuators and the structure via sensors and cable lines.
When the structure is subjected to an earthquake loading, the structural response at the top floor,
for example u(t) in the x-direction, is sent to the controller along with the earthquake excitations.
Optimal control forces from actuators are needed to minimize the structural displacement at the
top of the structure in every time step. Each optimal control force is converted to a corresponding
electric control signal E s (t) based on the properties of the actuator to be discussed later. Each
actuator is then excited by the control signal to generate a required control force.
Non-linear hysteretic effect
In order to represent the material non-linear and hysteretic behavior of a highrise building
structure, the restoring force of a structural element at time t, R(x, t), is defined by the following
equation proposed by Wen [42]:
R(x, t) = akx +(1a)kd y
(5)
where x is the displacement of the element, RL (x) = kx and RNL (v) = (1)kd y represent the
linear and non-linear parts of the restoring force, respectively, in which v represents the non-linear
displacement. The parameters k and k (<1) are the slope coefficients for linear and non-linear
parts of the loaddisplacement curve (Figure 3), respectively; and is a hysteretic variable in the
range 11, which is determined by solving the following non-linear differential equation [42]:

1
1+sgn(x)
v(t)
= x 1
||n
(6)
dy
2
where n is an integer number representing the yield exponent. It is used to control the sharpness of
the transition from the elastic part of the curve to the plastic region (noted by n in Figure 3). For
a bi-linear hysteretic representation, n = 1. The parameter d y is the displacement of the element at
the yield point when n = 1 (for the case of bi-linear hysteretic representation). The notation sgn is
the sign function. It is equal to 1 when its argument is positive and 1 otherwise. Equation (6) is
solved iteratively in each time step of the structural dynamic analysis.
Copyright q

DOI: 10.1002/nme
1050
Active actuator
system
xg (t)
Controller
(t)
u(t)
Sensor
Es(t)
Actuator
Sensor
Es(t)
Seismic
excitation xg (t)
Sensor
Figure 2. Building with an active control system in one direction. x g (t) = seismic acceleration excitation at time t, x(t) = structural displacement response, E s (t) = electric signal
to the actuator, f c (t) = control force from the actuator.
Equations (5) and (6) represent the non-linear stiffness hysteretic model of a structural element
where its restoring force is a function of a displacement variable x and a hysteretic variable v.
The restoring force vector for the entire structure is expressed in matrix form as
R(x, t) = KG X(t)+KM V(t)
(7)
where KG and KM are structural stiffness matrices. The stiffness matrix KG includes geometric
non-linearity terms, whereas stiffness matrix KM includes both geometric and material non-linearity
terms; X(t) and V(t) are the total displacement and non-linear displacement variable vectors at
time t (representing the second term in the right-hand side of Equation (5)), respectively.
Electrohydraulic actuator
The importance of the actuatorstructure coupling in active control of structures has been recognized
in the recent literature [8, 17, 28, 32, 43]. The linear hydraulic actuator is employed in this study for
active control of a building structure because of its ability to generate relatively large forces with a
relatively small response time in the order of a few milli-seconds and maintain the force with very
little power [8]. The level of force produced by a hydraulic actuator can be in the order of 1000 kN
(220 kilo-pounds) (http://www.nees.buffalo.edu/pdfs/244 Actuators.pdf). The magnitude of the
Copyright q

DOI: 10.1002/nme
1051
Force
fy
k
k
n
k
1
dy
x
Displacement
fy = kdy = yield force
Figure 3. Loaddisplacement curve for an element considering material

non-linearity and hysteretic behavior.
control force, F(t), is found by solving the following differential equation numerically in each
time step of the dynamic analysis [44]:
q(t) = Ap xp +
Cl
Vc
F(t)+
F(t)
Ap
2Ap
(8)
where Ap is the effective cross-sectional area of the piston; xp is the piston velocity; Cl is the
leakage coefficient; Vc is the volume of the cylinder; and is the compressibility coefficient of
the fluid. The displacement of the actuator/piston is assumed to be the same as the displacement
of the floor it is attached to.
State-space model
Considering material non-linearity and hysteretic behavior represented by Equation (5), geometrical
non-linear behavior of the structure, and the dynamics of the actuator represented by Equation (8),
the non-linear equation of the motion for a structure with an active control system subjected to
earthquake loading is expressed in the state-space form as follows:
= AZ(t)+BQc (t)+BFg (t)
Z(t)
(9)
T is a column vector of 4m state-space variables, in which
where Z(t) = [u(t), V(t), F (t), u(t)]

F (t) = Ic F(t). The 4m 4m matrix A is expressed by
0
0
0
I
0
0
0
D
(10)
A=
0
0
G1
G2
M1 KG
Copyright q
M1 KM
M1 Ic
M1 C

DOI: 10.1002/nme
1052
where D, KG , and KM are m m matrices incorporating the structural stiffness, geometrical and
material non-linearities. The terms in these matrices are defined as follows:

ki,effj
1+sgn(xi, j i, j )
n
|i, j |
Di, j =
1
(11)
fy
2
(K G )i, j = ki,effj
(12)
(K M )i, j = (1)ki,effj d yi, j
(13)
where the superscript eff on k denotes the effective lateral stiffness of the column element which
takes into account the geometrical non-linearity.
In Equation (10), 0 and I are the m m zero and identity matrices, respectively, and G1 and G2
are two m m diagonal matrices representing the characteristics of hydraulic actuators as follows:
2 C
x lx
VC x
G1 = diag 0
G2 = diag
0
0
2 y A2P y
VC y
,...,
2x A2P x
VC x
2 y Cly
VC y
2x A2P x
VC x
0
2 y Cly
VC y
2 C
x lx
VC x
,
.
.
.
,
0
0
0
0
0
2 y A2P y
VC y
0
(14)
L1
(15)
L1
The matrices Qc (t) and Fg (t) are 4m 1 actuator flow rate and external excitation vectors, respectively, expressed as follows:
01
01
01
01
(16)
Qc (t) =
, Fg (t) =
01
Ic q(t)
01
M0 Ig x g (t)
where q(t) is a 2L 1 vector representing the flow rate of hydraulic actuators [44] and
B = 01 01 01 M1 I1
(17)
is a 4m 1 matrix, in which 01 = [0 0 0]T and I1 = [1 1 1]T are m 1 zero and unit

column vectors, respectively.
The non-linear state-space equation represented by Equation (9) is solved numerically using the
fourth-order RungeKutta method with a proper integration time step (e.g. t = 0.01 s, a smaller
time step may be required for a larger structure). In order to obtain the mass, stiffness, and damping
Copyright q

DOI: 10.1002/nme
1053
coefficient matrices of the structure, building structures are modeled with finite elements in two
steps. In the first step, columns and beams are modeled as 3D frame elements with two end
nodes, each node having six DOFs (three displacements and three rotations). In the second step,
dynamic condensation is applied to reduce the structural model to one with rigid floor diaphragms
and no axial column deformations. The resulting structural model has three DOFs for each floor
(translations in x- and y-directions and a rotation about the vertical axis z passing through the
center of resistance, shown in Figure 1).
DYNAMIC FUZZY WNN EMULATOR

Recently, the authors developed a novel multi-paradigm dynamic time-delay fuzzy WNN model
for nonparametric identification of structures using the non-linear autoregression moving average
with exogenous inputs (NARMAX) approach [45]. The model integrates two intelligent computing
techniques (dynamic neural networks and fuzzy logic), a signal processing technique (wavelet
transform), and the chaos theory with the goal of improving the accuracy and adaptability of
nonparametric system identification. Furthermore, an adaptive LevenbergMarquardt-least-squares
(LM-LS) algorithm with a backtracking inexact linear search scheme has been developed for
training of the dynamic fuzzy WNN model [46]. The approach avoids the second-order differentiation required in the GaussNewton algorithm and overcomes the numerical instabilities encountered
in the steepest descent algorithm with improved learning convergence rate and high-computational
efficiency.
In order to control the response of a given structure effectively the structural response in
future time steps have to be estimated. This is necessary in order to determine the magnitude of the required control forces. In this study, the dynamic fuzzy WNN model developed
by the authors has been used as the neuroemulator to predict the non-linear structural response
in future time steps from the immediate past structural response and actuator dynamics. The
general discrete dynamic inputoutput mapping in the dynamic fuzzy WNN approach is expressed
as follows [45]:
D

M
D

Xk j ci j
f (Xk ) = wi
+ b j Xk j +d0 , k = 1, . . . , Na , a , (.) L 2 () (18)

a
ij
j
i=1
j=1
where f(Xk ) is the predicted structural response; (.) is the nonorthogonal Mexican hat wavelet
function used in this work [47]; X k j is the jth value in the kth input vector (or data point) Xk ,
and ci j is the jth value in the ith cluster of the multidimensional input vector obtained using the
fuzzy C-means clustering approach [48]. The parameter D is the dimension or the size of the input
vector in the NARMAX approach. The parameter M is the number of wavelets obtained using the
modified GramSchmidt algorithm [49] and the Akaikes final prediction error method (AFPE)
[50], which is also equal to the number of the fuzzy clusters as well as the number of wavelet nodes
used in the WNN model. The parameters ai j = 0 denote the frequency (or scale) corresponding
to the multidimensional input vector; wi represents the ith wavelet coefficient linking the hidden
node to the output; b j is the weight of the link of the jth input to the output (for the linear part
of the neural network, input is connected directly to the output); d0 is a bias term, and is the
set of real numbers. The notation L 2 () represents the square summable constructed state-space
vectors. The parameter Na is the number of input vectors.
Copyright q

DOI: 10.1002/nme
1054
Since the earthquake can occur in any horizontal direction two fuzzy WNN-based emulators,
x- and y-neuroemulators, are created to predict the corresponding structural displacement responses
in the x- and y-directions. Details of how to construct the dynamic fuzzy WNN model using
a given time series are presented in Adeli and Jiang [45]. The procedures to construct the
two neuroemulators are similar. The response data in the x-direction are used to construct the
x-neuroemulator and those in the y-direction are used for the y-neuroemulator. The resulting
numbers of the hidden or wavelet nodes in the hidden layer of the two neuroemulators are in
general different.
TRAINING OF THE FUZZY WNN NEUROEMULATOR

Bani-Hani and Ghaboussi [17] investigate two neuroemulators using the BP neural network algorithm: linearly trained neuroemulator and non-linearly trained neuroemulator. They show that the
emulator trained using structural non-linear response data can predict structural responses more
accurately than that trained using structural linear response data, especially when the structure is
subjected to strong earthquake excitations. The neuroemulator presented in this paper is a nonphysically based model (or black box), which does not explicitly involve any physical parameters
of the structure, in the context of both training and testing. The neuroemulator is trained to map
non-linear relationships between input and output data to represent the physical model of the
corresponding structure; then the trained neuroemulator is used to predict the structural response
under various input excitations.
In the current study, the dynamic fuzzy WNN neuroemulator is trained using the structural
displacement response data generated from the non-linear structural analysis taking into account
both geometrical and material non-linearities as described previously. The training data are generated from the non-linear analyses of the structure subjected to different earthquake loadings.
Figure 4 shows the training diagram of the dynamic fuzzy WNN model using the adaptive LM-LS
learning algorithm for active non-linear control of structures. The training is done in three steps,
identified by numbers (1)(3) in Figure 4 as follows:
(1) The inputoutput data sets, Xk and yk (k = 1, 2, . . . , Na ), are formed from structural
responses, actuator forces, and earthquake excitations, where the kth input state-space
vector (time t during the earthquake excitation), Xk , is constructed as follows:
Xk = [xk , fk , ak ]T = xk , xk+1 , . . . , xk+x , Fk , Fk+1 , . . . , Fk+c , x gk , x gk+1 , . . . , x gk+g T
(19)
where xk , Fk , and x gk are the kth structural displacement response, actuator/control force,
and earthquake excitation, respectively. The parameters x , c , and g are time delays for
the three variables representing dimensions of the inputs (number of previous discrete time
steps used for any input time series). Their optimal dimensions are obtained by the false
nearest neighbor (FNN) method, as discussed in Jiang and Adeli [51]. The number of data
sets Na is obtained by
Na = N max(x , c , g )
(20)
where N is the total number of data points used to train the system.
(2) The dynamic fuzzy WNN model with a recurrent feedback topology is created as described
in Adeli and Jiang [45].
Copyright q

DOI: 10.1002/nme
y k 1
y
2) Dynamic
fuzzy WNN
model
Sensor
F (t )
Feedback
1055
C-mean
Controller
x(t)
..
x g (t )
y k = f ( X k )
1) Construct
State space
Structural
system
E
yk
X
Update
parameters
3) Training model
Adaptive LM-LS
algorithm
Seismic excitation
Converge?
No
Yes
(t)
Stop
Figure 4. Training diagram of the dynamic fuzzy WNN model using the adaptive LM-LS learning algorithm for active non-linear control of structures. F(t) = actuator/control force at time t, x g (t) = ground
acceleration, y(t) = measured structural response at time t, yk = kth measured structural response
(k represents time step t), X = [X1 X2 Xk ]T = input state space vector for all structural response data,
f(Xk ) = fuzzy WNN mapping function represented by Equation (18), yk = f(Xk ) = kth computed output,
y k1 = feedback structural response, E = error function.
(3) The dynamic fuzzy WNN model is trained using the adaptive LM-LS algorithm and the
constructed inputoutput data sets [46].
The trained neuroemulator is tested using the data generated from both non-linear and linear
structural analyses. The ability of the trained model to predict both linear and non-linear responses
of the structure is investigated.
FINDING OPTIMAL CONTROL FORCES

A floating-point genetic algorithm is developed for finding the optimal control forces for active
linear or non-linear control of highrise building structures in tandem with the dynamic fuzzy
wavelet neuroemulator presented in this paper. The corresponding control signals are sent to the
actuators that generate the control forces to be applied on the structure. Details of the genetic
control algorithm and how to implement the control methodology are presented in the companion
paper [40].
Copyright q

DOI: 10.1002/nme
1056
NUMERICAL VALIDATION OF THE DYNAMIC FUZZY WAVELET NEUROEMULATOR

In order to demonstrate that the new dynamic fuzzy wavelet neuroemulator model can accurately
predict the dynamic responses of structures with irregularities, two 3D multistory steel building
structures, a 12-story structure with vertical setbacks (Example 1), and an 8-story structure with
plan irregularity (Example 2) are studied. Kim and Adeli [41] used the same examples in their
study of hybrid control of irregular highrise building structures assuming linear behavior. They
report that a 2D dynamic analysis neglecting the effect of coupling between lateral and torsional
motions underestimates the maximum response of the structure up to 4% for Example 1 and 7%
for Example 2.
The non-linear dynamic responses of the two example structures subjected to earthquake excitations are obtained by solving their dynamic equation of motion represented by Equation (6).
Both geometric and material non-linearities are considered in the structural analysis. A steel yield
stress of f y = 36 ksi (248.3 N/mm2 in Equation (5)), a yield ratio of = 0.1 and a yield exponent
of n = 2 (Equation (5)), and a damping ratio of 2% are used for both structures.
The original Chichi earthquake and 200% Chichi earthquake record of Richter magnitude
7.3, which occurred at Foothills, Western Taiwan, on 21 September 1999, are applied to the
structures. The structure is assumed to behave linearly when the difference of the maximum
displacement obtained from the non-linear and linear analyses is within 1% of the linear maximum
displacement. On the basis of this definition, under the Chichi earthquake loading, the 12-story
structure (Example 1) behaves non-linearly while the 8-story structure (Example 2) behaves linearly.
When the 200% Chichi earthquake loading is applied, both structures behave non-linearly.
The displacement responses of the uncontrolled structure obtained from 200% Chichi earthquake representing the structural non-linear behavior are used to train the neuroemulator, while
the displacement responses from the original Chichi earthquake excitation are used to test the
neuroemulator. Displacement responses are obtained at each floor in the x and y horizontal directions at increments of 0.01 s over a period of 40 s, equal to the duration of the Chichi earthquake
excitation. Thus, the number of sample data, N in Equation (20), is 4000.
The non-linear control model is trained using the results obtained when the structure is subjected
to (a) earthquake excitation (uncontrolled structure) and (b) both earthquake excitation and actuator
forces simultaneously. This will enhance the adaptability of the fuzzy wavelet neuroemulator
because for some cases no control forces may be necessary, for example, when earthquake excitation
is small.
Example 1: Twelve-story irregular steel building
Example structure and its response. This example is an irregular 12-story 3D steel building
structure with vertical setbacks as shown in Figure 5. This structure was created by Adeli and
Saleh [13, 14] for the study of structural control and used by Jiang and Adeli [46] for non-linear
system identification of structures and Kim and Adeli [41] for hybrid control of the irregular
structure. It consists of 148 members and 77 nodes. In earlier papers [14, 46], the structure was
modeled as a 3D space frame with 462 DOFs. In this article, floor diaphragms are assumed
to be rigid and axial column deformations are neglected. The rigid-floor structural model has
36 DOFs.
The members of the structure are W shapes as indicated in Figure 5. The structure is subjected
to the combination of uniformly distributed floor dead and live loads of 60 psf (2.88 kPa) and 50 psf
(2.38 kPa), respectively, and the Chichi earthquake records. The Chichi earthquake records were
Copyright q

DOI: 10.1002/nme
1057
W14 61
z A
Section I
W21 50
W21 57
W21 50
(b)
W14 90
12 4.5m
W21 57
W21 50
W21 50
W21 50
W14 68
Section II
W21 50
2 5.5m
Ground
excitation
x
(a)
=-13
Beam/column
y
(c)
Actuator location
Figure 5. Twelve-story steel building with vertical irregularity (Example 1): (a) perspective
view; (b) plan of section I; and (c) plan of section II.
used as the representative of one of the most destructive earthquakes of the past few decades for
illustration. The modal analyses of this structure conducted by Kim and Adeli [41] demonstrate
that the coupling effect of lateral and torsional vibrations is most significant when the earthquake
excitation is applied in the direction with = 13 (Figure 5). Therefore, this is the angle used
in this research to perform non-linear dynamic analyses of the structure for training and testing
the fuzzy wavelet neuroemulator. Figure 6 shows the x and y displacements of joint A at the top
floor of the structure (identified in Figure 5) when the structure is subjected to the original Chichi
and 200% of Chichi earthquakes in the direction with = 13 .
Actuators. Four double-acting actuators with the same properties are installed in every floor along
the exterior envelope of the structure as indicated with dashed lines in Figures 5(b) and (c). The
properties of an actuator used in this study are based on data provided by one manufacturer
(http://www.mts.com/vehicles/testline/pdfs/100-016-993.pdf) and are summarized in Table I.
Copyright q

DOI: 10.1002/nme
1058
Acceleration (g)
2
1.5
1
0.5
0
-0.5
-1
(a) -1.5
Displacement (m)
0.5
0
-0.5
-1
10
15
20
t (s)
25
30
35
40
10
15
20
t (s)
25
30
35
40
(b)
Displacement (m)
0.3
0.2
0.1
0
-0.1
-0.2
(c)
Original earthquake
200% of original earthquake
Figure 6. Displacements of joint A at the top floor of the structure (identified in Figure 7) when the
structure is subjected to the original Chichi and 200% Chichi earthquakes in the direction with = 13 :
(a) 200% Chichi earthquake, Taiwan, 21 September 1999; (b) displacements of joint A in the x-direction;
and (c) displacements of joint A in the y-direction.
Table I. Properties of the actuator.

Variables
Ap
Vc
Cl
qmax
Fmax
g
Copyright q
Description
Value
Effective piston area

Chamber volume
Leakage coefficient
Compressibility coefficient
Maximum flow rate
Maximum actuator force
Servovalve time constant
Servovalve constant
3.368103 m2
1.01103 m3
0.11010 m5 /(N s)
2.11010 N/m2
2.0103 m3 /s
68 kN
0.15 s
2.1104 m3 /s/volt

DOI: 10.1002/nme
1059
x104
6
F(t) (N)
3
0
-3
-6
0
10
15
20
t(s)
25
30
35
40
Figure 7. Randomly generated control forces in the x-direction over the same earthquake duration of 40 s
for the actuator with properties shown in Table I.
For training the non-linear control model, two sets of control forces are randomly generated in the
x- and y-directions with a maximum actuator force of 68 kN (Table I) and a uniform distribution
in the interval of (68, 68 kN). Figure 7 shows the randomly generated control forces in the
x-direction over the same earthquake duration of 40 s.
Constructing fuzzy wavelet neuroemulators. Three sets of time series data are needed to construct
each one of the two fuzzy WNN neuroemulators in the x- and y-directions: the structural
displacement response at joint A (Figure 5(a)), earthquake excitation, and control force. Each
set of time series consists of 4000 data points. Optimum time-delay steps of x = 2, g = 5, and
u = 8 are obtained for the displacement response at joint A, the earthquake excitation, and the
control force using the FNN method, respectively. The same numbers of time steps are obtained
for the data sets in both x- and y-directions. The number of input vectors created is thus equal to
Na = 4000max{2, 5, 8} = 3992 (Equation (20)).
A five-level wavelet decomposition was performed on the 3992 sets of input vectors with the size
of x +u +g = 15 (Equation (18)) using the Mexican hat wavelet function. The empty wavelets
whose supports do not contain any data are eliminated, resulting in 109 and 105 nonempty wavelets
for x- and y-direction vectors, respectively. On the basis of the AFPE criterion and using the
modified GramSchmidt algorithm [52], it is concluded that three out of the 109 and four out
of the 105 nonempty wavelets are sufficient to construct x- and y-fuzzy WNN neuroemulators,
respectively. As such, there are three WNN nodes in the hidden layer of the x-neuroemulator for
predicting the x-directional displacement response and four WNN nodes in the hidden layer of the
y-neuroemulator for predicting the y-directional displacement response. For a discussion on the
concept of empty wavelets and how the necessary number of nonempty wavelets is chosen, refer
to Jiang and Adeli [52].
Training fuzzy wavelet neuroemulators. Two sets of data are used to train each neuroemulator,
one set at a time, denoted as Set1 and Set2. In Set1, original Chichi earthquake excitations
during the previous g = 5 time intervals and the resulting structural displacement responses of
joints A (Figure 5(a)) during the previous x = 2 time intervals are used as inputs. The structural
responses in the x-direction (Figure 6(b)) are used for training the x-neuroemulator and those in the
Copyright q

DOI: 10.1002/nme
1060
y-direction (Figure 6(c)) for the y-neuroemulator. In Set2, 200% of Chichi earthquake excitations
(scaled) during the previous g = 5 time intervals, the randomly generated control forces during
the previous u = 8 time intervals, and the structural displacement responses of joints A during the
previous x = 2 time intervals subjected to combined earthquake and control loadings are used as
inputs. The control forces in the x-direction (Figure 7) and their corresponding responses are used
for training the x-neuroemulator and those in the y-direction for training the y-neuroemulator. In
both sets of training, the current displacement response of joint A is the output of the dynamic
fuzzy wavelet neuroemulators. Set1 data are used to illustrate the effectiveness of the proposed
neuroemulator in predicting linear response of the structure, while Set2 data are used to illustrate
its effectiveness in predicting non-linear response of the structure.
The training of the model using the adaptive LM-LS learning algorithm converges very fast after
only six iterations for training the x-neuroemulator and eight iterations for the y-neuroemulator
using each of the two sets of data. The model is implemented in a combination of C++ programming language and MATLAB 6.1 [53] on a Windows XP Professional platform and a 1.5 GHz
Intel Pentium 4 processor. The CPU time for training the x-neuroemulator using each set of data
is only 53 s and for training the y-neuroemulator, only 55 s.
Testing fuzzy wavelet neuroemulators. The performance of the trained neuroemulators is evaluated
by comparing their predicted displacement response with the finite element simulation results
obtained directly by Equation (9). The evaluation is conducted in both time and frequency domains
for non-linear response of the structure subjected to 100% of Chichi earthquake. Each of the
trained x- and y-neuroemulators is tested using two sets of data defined previously for 100% of
Chichi earthquake. Figure 8 shows the predicted and simulated structural displacement responses
of joint A of the 12-story structure in the x- and y-directions under the combined Chichi earthquake
excitation and actuator forces. The relative root mean square (RRMS) errors between the simulated
and predicted results for the four test cases (two in the x and two in the y-direction) are summarized
in Table II. The RRMS errors of the displacement responses for all four sets of test data are less
than 0.3, which is quite small.
The pseudospectra method, a power density spectrum method, proposed by authors for damage
detection of highrise buildings [54] is employed in this study to evaluate the accuracy of the trained
neuroemulators in the frequency domain. The multiple signal classification (MUSIC) method is
employed to compute the pseudospectrum from the structural response time series. Figure 9 shows
a comparison of pseudospectra of the predicted and simulated displacement responses of joint A
of the 12-story structure in the x- and y-directions under the combined earthquake and actuator
loadings. The pseudospectra of the predicted and simulated structural responses match quite well
for both x- and y-directions, thus demonstrating that the fuzzy WNN neuroemulators provide
accurate prediction of structural displacement responses.
Example 2: Eight-story irregular steel building
Example structure and its response. This example structure is an 8-story 3D steel building structure
with a plan irregularity and a height of 36 m shown in Figure 10. This structure was created by
Kim and Adeli [39] for the study of hybrid control of 3D irregular buildings. It consists of 208
members and 99 nodes. The members of the structures are various W shapes as described in Kim
and Adeli [39]. Floor diaphragms are assumed to be rigid and axial column deformations are
neglected. The resulting structural model has 24 DOFs.
Copyright q

DOI: 10.1002/nme
1061
Displacement (m)
1
0.5
0
-0.5
-1
10
15
20
t(s)
25
30
35
40
10
15
20
t(s)
25
30
35
40
(a)
1.5
1
0.5
0
-0.5
-1
-1.5
(b)
Simulated data
Predicted by fuzzy
wavelet neuroemulator
Figure 8. Predicted and simulated displacement responses of joint A of a 12-story structure (Example 1)
subjected to combined Chichi earthquake excitation and actuator forces: (a) displacements of joint A in
the x-direction and (b) displacements of joint A in the y-direction.
Table II. Relative root mean square (RRMS) error of four test cases in Example 1.
Loading type
Earthquake
excitation
Combined earthquake excitation

and actuator forces
0.29
0.23
0.15
0.24
x-direction
y-direction
The structure is subjected to the combination of uniformly distributed floor dead and live loads
of 100 psf (4.78 kPa) and 70 psf (3.35 kPa), respectively, and the Chichi earthquake records. The
modal analyses of this structure conducted by Kim and Adeli [39] demonstrate that the coupling
effect of lateral and torsional vibrations is most significant when the earthquake excitation is
applied in the direction with = 83.4 (Figure 10). Therefore, this is the angle used in this research
to perform non-linear dynamic analyses of the structure for training and testing the new fuzzy
wavelet neuroemulator.
Actuators. Four double-acting actuators with the same properties are installed in every floor as
indicated with dashed lines in Figure 10(b). Their properties are the same as those employed
in Example 1. The control forces are randomly generated and used similar to that explained for
Example 1.
Copyright q

DOI: 10.1002/nme
1062
Magnitude (dB)
150
100
50
0
-50
/10
/5 3/10 2/5 /2 3/5 7/10 4/5 9/10

Frequency (Hz)
/10
/5 3/10 2/5 /2 3/5 7/10 4/5 9/10

Frequency (Hz)
(a)
Magnitude (dB)
150
100
50
0
-50
(b)
Simulated data
Predicted by fuzzy
Figure 9. Comparison of pseudospectra of predicted and simulated displacement responses of joint A

of 12-story structure (Example 1) under combined earthquake and actuator loadings: (a) comparison of
displacements in the x-direction and (b) comparison of displacements in the y-direction.
8 4.5m = 36m
y
2 6m = 12m
3 6m = 18m
(b)
x
Beam
=83.4 0
Ground excitation
Actuator
location
(a)
Figure 10. Eight-story steel building with plan irregularity (Example 2): (a) perspective view and (b) plan.
Copyright q

DOI: 10.1002/nme
1063
Constructing fuzzy wavelet neuroemulators. Following the procedure for constructing the neuroemulators explained for Example 1, optimum time-delay steps of x = 2, g = 5, and u = 14 are
obtained for the displacement response at joint A, the earthquake excitation, and the control
force using the FNN method, respectively. The number of input vectors created is equal to
Na = 4000max{2, 5, 14} = 3986 (Equation (19)). The 3986 sets of input vectors are used to
construct the neuroemulators. The resulting x-neuroemulator has 21 input nodes in the input layer
(x +u +g = 21 in Equation (18)) and two hidden/wavelet nodes in the hidden layer, whereas the
resulting y-neuroemulator has 21 input nodes in the input layer and three hidden/wavelet nodes
in the hidden layer.
Training fuzzy wavelet neuroemulators. Two sets of data, denoted by Set1 and Set2, similar to
those explained for Example 1, are used to train each neuroemulator, one set at a time. The training
of the model using the adaptive LM-LS learning algorithm converges after only four iterations for
training the x-neuroemulator and six iterations for the y-neuroemulator using each of the two sets
of data.
Testing fuzzy wavelet neuroemulators. Testing is performed similar to that for Example 1 under
100% of Chichi earthquake. This structure, however, behaves linearly under this earthquake loading
as mentioned earlier. Figure 11 shows the predicted and simulated displacement responses of
Displacement (m)
1
0.5
0
-0.5
-1
10
15
20
t(s)
25
30
35
40
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
0
10
15
20
t(s)
25
30
35
40
Displacement (m)
(a)
(b)
Simulated data
Predicted by fuzzy
Figure 11. Predicted and simulated displacement responses of joint A of an 8-story building (Example 2)
subjected to combined Chichi earthquake excitation and actuator forces: (a) displacements of joint A in
the x-direction and (b) displacements of joint A in the y-direction.
Copyright q

DOI: 10.1002/nme
1064
Table III. Relative root mean square (RRMS) error of four test cases in Example 2.
Loading type
Earthquake
excitation
Combined earthquake excitation

and actuator forces
0.22
0.20
0.14
0.19
x-direction
y-direction
joint A of the structure in the x- and y-directions under the combined Chichi earthquake excitation
and actuator forces. The RRMS errors between the simulated and predicted results for the four test
cases (two in the x and two in the y-direction) are summarized in Table III. The RRMS errors of
the displacement responses for all four sets of test data are less than 0.25 which is quite small. The
performance of the trained neuroemulators is also evaluated using the pseudospectra method in the
frequency domain for the Chichi earthquake. The pseudospectra of the predicted and simulated
structural responses match quite well for both x- and y-directions.
CONCLUDING REMARKS
An effective model has been developed for active non-linear control of large 3D structures. Both
geometric and material non-linearities are included in the control formulation. Coupling between
lateral and torsional motions of the structure as well as the actuator dynamics are taken into account
in the control model.
A dynamic fuzzy wavelet neuroemulator was presented for predicting the structural response
in future time steps. It is demonstrated that the dynamic fuzzy wavelet neuroemulator provides
accurate prediction of structural displacement responses in both linear and non-linear ranges. In
the companion paper [40], a floating-point genetic algorithm is presented for finding the optimal
control forces needed for active non-linear control of building structures.
REFERENCES
1. Masri SF, Bekey GA, Caughey TK. On-line control of nonlinear flexible structures. Journal of Applied Mechanics
1982; 49(12):871884.
2. Stein G, Athans M. The LQG/LTR procedure for multivariable feedback control design. IEEE Transactions on
Automatic Control 1987; AC32(2):105114.
3. Miller RK, Masri SF, Dehghanyar T, Caughey TK. Active vibration control of large civil structures. Journal of
Engineering Mechanics (ASCE) 1988; 114(9):15421570.
4. Yang JN, Long FX, Wong D. Optimal control of nonlinear structures. Journal of Applied Mechanics 1988;
55(4):931938.
5. Yang JN, Danielians A, Liu SC. Hybrid control of nonlinear and hysteretic systems I&II. Journal of Engineering
Mechanics (ASCE) 1992; 118:14231439, 14411457.
6. Soong TT. Active Structural Control: Theory and Practice. Longmans: New York, 1990.
7. Chang CC, Yang HTY. Control of building using active tuned mass dampers. Journal of Engineering Mechanics
(ASCE) 1995; 121(3):355366.
8. Nikzad K, Ghaboussi J, Paul SL. Actuator dynamics and delay compensation using neurocontrollers. Journal of
9. Tomasula DP, Spencer BF, Sain MK. Nonlinear control strategies for limiting dynamic response extremes. Journal
of Engineering Mechanics (ASCE) 1996; 122(3):218229.
10. Thomson M, Schooling SP, Soufian M. The practical application of a nonlinear identification methodology.
Control Engineering Practice 1996; 4(3):295306.
Copyright q

DOI: 10.1002/nme
1065
11. Housner GW, Bergman LA, Caughey TK, Chassiakos AG, Claus RO, Masri SF, Skelton RE, Soong TT, Spencer
BF, Yao JTP. Structural control: past, present, and future. Journal of Engineering Mechanics (ASCE) 1997;
123(9):897971.
12. Adeli H, Saleh A. Optimal control of adaptive/smart bridge structures. Journal of Structural Engineering (ASCE)
1997; 123(2):218226.
13. Adeli H, Saleh A. Control, Optimization, and Smart StructuresHigh-performance Bridges and Buildings of the
Future. Wiley: New York, 1999.
14. Saleh A, Adeli H. Optimal control of adaptive/smart multistory building structures. Computer-Aided Civil and
Infrastructure Engineering 1998; 13(6):389403.
15. Lu LT, Chiang WL, Tang JP. LQG/LTR control methodology in active structural control. Journal of Engineering
Mechanics (ASCE) 1998; 124(4):446454.
16. Bani-Hani K, Ghaboussi J. Neural networks for structural control of a benchmark problem, active tendon system.
Earthquake Engineering and Structural Dynamics 1998; 27(11):12251245.
17. Bani-Hani K, Ghaboussi J. Nonlinear structural control using neural networks. Journal of Engineering Mechanics
(ASCE) 1998; 124(3):319327.
18. Hung SL, Kao CY, Lee JC. Active pulse structural control using artificial neural networks. Journal of Engineering
Mechanics (ASCE) 2000; 126(8):839849.
19. Connor JJ. Introduction to Structural Motion Control. Prentice-Hall: Upper Saddle River, NJ, 2003.
20. Adeli H, Kim H. Wavelet-hybrid feedback-least mean square algorithm for robust control of structures. Journal
of Structural Engineering (ASCE) 2004; 130(2):128137.
21. Kim H, Adeli H. Hybrid feedback-least mean square algorithm for structural control. Journal of Structural
Engineering (ASCE) 2004; 130(2):120127.
22. Edwards C, Spurgeon S. Sliding Mode Control: Theory and Applications. Taylor & Francis, CRC Press: London,
Boca Raton, FL, 1998.
23. Agrawal AK, Yang JN. Optimal polynomial control of seismically excited linear structures. Journal of Engineering
Mechanics (ASCE) 1996; 122(8):753761.
24. Adeli H, Hung SL. Machine LearningNeural Networks, Genetic Algorithms, and Fuzzy Sets. Wiley: New York,
1995.
25. Adeli H, Park HS. Neurocomputing for Design Automation. CRC Press: Boca Raton, FL, 1998.
26. Adeli H. Neural networks in civil engineering: 19892000. Computer-Aided Civil and Infrastructure Engineering
2001; 16(2):126142.
27. Chen HM, Tsai KH, Qi GZ, Yang JCS, Amini F. Neural network for structure control. Journal of Computing
in Civil Engineering (ASCE) 1995; 9(2):168176.
28. Ghaboussi J, Joghataie A. Active control of structures using neural networks. Journal of Engineering Mechanics
(ASCE) 1995; 121(4):555567.
29. Kim JT, Jung HJ, Lee IW. Optimal structural control using neural networks. Journal of Engineering Mechanics
(ASCE) 2000; 126(2):201205.
30. Kim DH, Seo SH, Lee IW. Optimal neurocontroller for nonlinear benchmark structure. Journal of Engineering
Mechanics (ASCE) 2004; 130(4):424429.
31. Brown AS, Yang HTY. Neural networks for multiobjective adaptive structural control. Journal of Structural
32. Kim DH, Lee IW. Neurocontrol of seismically excited steel structures through sensitivity evaluation scheme.
Earthquake Engineering and Structural Dynamics 2001; 30(9):13611377.
33. Fur LS, Yang HTY, Ankireddi S. Vibration control of tall building buildings under seismic and wind loads.
Journal of Structural Engineering (ASCE) 1996; 122(8):948957.
34. Al-Dawod M, Smali B, Kwok K, Naghdy F. Fuzzy controller for seismically excited nonlinear buildings. Journal
of Engineering Mechanics (ASCE) 2004; 130(4):407415.
35. Ohtori Y, Christenson RE, Spencer BF, Dyke SJ. Benchmark control problems for seismically excited nonlinear
buildings. Journal of Engineering Mechanics (ASCE) 2004; 130(4):366385.
36. Yoshida O, Dyke SJ. Response control of full-scale irregular buildings using magnetorheological dampers. Journal
37. Tan P, Dyke SJ. Integrated device placement and control design in civil structures using genetic algorithms.
Journal of Structural Engineering (ASCE) 2005; 131(10):14891496.
38. Kim H, Adeli H. Hybrid control of smart structures using a novel wavelet-based algorithm. Computer-Aided
Civil and Infrastructure Engineering 2005; 20(1):722.
Copyright q

DOI: 10.1002/nme
1066
39. Kim H, Adeli H. Wavelet hybrid feedback-LMS algorithm for robust control of cable-stayed bridges. Journal of
Bridge Engineering (ASCE) 2005; 10(2):116123.
40. Jiang XM, Adeli H. Neuro-genetic algorithm for nonlinear active control of highrise buildings. International
Journal for Numerical Methods in Engineering 2007; DOI: 10.1002/nme.2274.
41. Kim H, Adeli H. Hybrid control of irregular steel highrise building structures under seismic excitations.
International Journal for Numerical Methods in Engineering 2005; 63(12):17571774.
42. Wen YK. Method for random vibration for hysteretic structures. Journal of the Engineering Mechanics Division
(ASCE) 1976; 102(EM2):249263.
43. Dyke SJ, Spencer BF, Sain MK. Role of controlstructure interaction in protective system design. Journal of
44. Desilva CW. Control Sensors and Actuators. Prentice-Hall: Englewood Cliffs, NJ, 1989.
45. Adeli H, Jiang XM. Dynamic fuzzy wavelet neural network model for structural system identification. Journal
46. Jiang XM, Adeli H. Dynamic wavelet neural network for nonlinear identification of highrise buildings. ComputerAided Civil and Infrastructure Engineering 2005; 20(5):316330.
47. Zhou Z, Adeli H. Time-frequency signal analysis of earthquake records using Mexican hat wavelets. ComputerAided Civil and Infrastructure Engineering 2003; 18(5):379389.
48. Bezdek JC. Pattern Recognition with Fuzzy Objective Function Algorithms. Plenum: New York, 1981.
49. Zhang Q. Using wavelet network in nonparametric estimate. IEEE Transactions on Neural Networks 1997;
8(2):227236.
50. Ljung L. System IdentificationTheory for the User (2nd edn). Prentice-Hall: Upper Saddle River, NJ, 1999.
51. Jiang XM, Adeli H. Fuzzy clustering approach for accurate embedding dimension identification in chaotic time
series. Integrated Computer-Aided Engineering 2003; 10(3):287302.
52. Jiang XM, Adeli H. Dynamic wavelet neural network model for traffic flow forecasting. Journal of Transportation
53. Matlab. The Language of Technical Computing. Mathworks Inc.: Natick, MA, 2001.
54. Adeli H, Jiang XM. Psuedospectra, MUSIC, and dynamic fuzzy wavelet neural network for damage detection
of highrise building using sensor data. International Journal for Numerical Methods in Engineering 2007;
71(5):606629.
Copyright q

DOI: 10.1002/nme

Dynamic Fuzzy Wavelet Neuroemulator For Non-Linear Control of Irregular Buildins Structures

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dynamic Fuzzy Wavelet Neuroemulator For Non-Linear Control of Irregular Buildins Structures

Uploaded by

Copyright:

Available Formats

INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING

Int. J. Numer. Meth. Engng 2008; 74:10451066