Available online at www.sciencedirect.

com

Journal of the Franklin Institute 360 (2023) 10536–10563

www.elsevier.com/locate/jfranklin

A new seismic control framework of optimal PIλDμ

controller series with fuzzy PD controller including
soil-structure interaction
Sadegh Etedali a,∗, Abbas-Ali Zamani b, Morteza Akbari c,
Mohammad Seifi c
a Department of Civil Engineering, Birjand University of Technology, P.O. Box 97175-569, Birjand, Iran
b Department of Electrical Engineering, Technical and Vocational University (TVU), Tehran, Iran
c Independent researchers, Birjand, Iran

Received 30 April 2022; received in revised form 20 April 2023; accepted 6 August 2023
Available online 9 August 2023

Abstract
A new framework of optimal fractional order proportional-derivative-integral (FOPID) controller se-
ries with fuzzy proportional-derivative (PD) controller, namely OFPD-FOPID controller, is proposed in
this study for seismic control of structures equipped with active tuned mass damper (ATMD). Three
controllers including optimal PID, optimal FOPID, and fuzzy PID (FPID) controllers are also imple-
mented for comparison purposes. Simulation results carried out on a 15-story building show the FOPID
controller than the PID and FPID controllers can remarkably reduce the maximum floor displacement,
but they represent a poor performance in mitigation of the maximum floor acceleration in different
soil conditions, while the proposed OFPD-FOPID controller tracking the amount of the maximum floor
acceleration to estimate the optimal control force of the actuator can provide superior performance. An
average reduction of 41%, 45%, and 33% in the maximum floor displacement; 36%, 33%, and 20%
in the maximum inter-story drift are given by FOPID in the dense, medium, and soft soils, while it
results in an increase of 45%, 52% and 24% in the maximum floor acceleration. Similarly, the proposed
OFPD-FOPID controller represents an average reduction of 52%, 55%, and 45% in the maximum floor
displacement; 42%, 44%, and 28% in the maximum inter-story drift in the dense, medium, and soft
soils, while it also slightly reduces the maximum floor acceleration of the studied structure located on
different soil conditions.
© 2023 The Franklin Institute. Published by Elsevier Inc. All rights reserved.

https://doi.org/10.1016/j.jfranklin.2023.08.003
0016-0032/© 2023 The Franklin Institute. Published by Elsevier Inc. All rights reserved.
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

1. Introduction

For a long time ago, researchers have paid close attention to mitigating the dynamic re-
sponses of civil structures to external conditions such as earthquakes and wind loads. Devel-
oping and study of vibration control devices including passive, active, semi-active, and hybrid
control devices for civil engineering structures is an admirable effort toward this aim. The
control devices in passive control methods do not apply any external forces and instead reduce
the vibration responses through input energy dissipation or shifting the dominant frequency
of the structure. Unlike the passive control strategy, the active control method applied an
external control force to the structure. The control force is usually a function of the measured
response of the structure and it is applied through an actuator to prevent the balance of the
structure from being disturbed and it to collapse. The structural responses are measured by
sensors and corresponding them, a suitable control algorithm commands the actuator to apply
the control force. Hence, the active control strategy is more complex than the passive control
because it requires the installation of sensors, actuators, and a control algorithm. However,
the active control strategy represents a superior performance in different magnitudes and a
high-frequency range of dynamic loads such as strong loads and earthquakes.
TMD is a well-known passive energy-absorbing device combined with a mass, a spring,
and a viscous damper. The TMD frequency is often tuned with the fundamental frequency of
the main structure. The performance of TMD for seismic-excited structures is well studied in
the literature [1–4]. Because the SSI affects the dynamic characteristics of a structure, some
researchers focus on the optimal design of the TMD parameters for seismic-excited structures
taking into account SSI effects [5–7]. Farshidianfar and Sohaili [5] optimized the TMD pa-
rameters using the ant colony algorithm for tall structures, taking into account the interaction
of the soil and the structure. By using the data of the Kobe and Tabas earthquakes and eval-
uating the results for a 40-story structure, they concluded that the type of soil significantly
affects the optimized parameters of TMD and the time response of structures. Khatibinia et al.
[6] proposed a multi-objective optimization of tuned mass dampers considering SSI effects.
They concluded that the SSI effects have a significant influence on the optimum parameters
of TMD. Using a multi-objective particle swarm algorithm, Etedali et al. [7] conducted a
parametric study on the optimal TMD and friction-tuned mass damper (FTMD) parameters in
a 3-story structure considering SSI effects. The simulation results also show that the FTMD
provides better performance in reducing the maximum top-floor displacement and acceleration
of the building in different soil types. TMD is an effective and practical system for control-
ling structural vibrations during some earthquakes; however, the efficiency of the TMD for
the pulse-like near-field ground motion has not been determined with certainty [8]. To en-
hance the performance of the conventional TMD, some modified versions of the TMD such as
tuned-viscous-mass-damper (TVMD), tuned-mass-damper-inerter (TMDI), and tuned-inerter-
damper (TID) [9–13]. Taflanidis et al. [12] investigated a multi-objective optimal design of
tuned-viscous-mass- TVMD, TMDI, and TID for earthquake protection of multistory building
structures. Liu et al. [13] provided a retrospective viewpoint and an update on the inerter’s
development for vibration isolation in various fields. They concluded, which highlights the
exigence of inerter research, prevalent challenges, and future research orientations.

∗ Corresponding author at: Department of Civil Engineering, Birjand University of Technology, Birjand, Iran.
E-mail address: [email protected] (S. Etedali).

10537
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

To overcome this drawback, the active form of the TMD namely ATMD has been suggested
with the aim of adaptability in earthquakes with different intensities and frequency content.
In an ATMD, the control force is applied to a TMD by an actuator based on the measured
responses of the structure using sensors, and the reaction of the control force is applied to the
structure to mitigate the seismic responses of the structure. For a successful implementation
of the ATMD in seismic-excited structures, a suitable control algorithm should be tuned to the
desired control force in real-time [14]. Gao et al. [15] focused on a single-degree-of-freedom
structure with an ATMD and suggest a hybrid learning control strategy to suppress vibrations
caused by unknown time-varying disturbances. Ignoring the SSI effects may result in non-
conservative and inaccurate results of the structures equipped with ATMD during earthquakes.
Consequently, some researchers recently addressed this issue in their studies [14,16-18].
The PID controller family as simple and practical control algorithms in the industry have
been attracted by researchers in the area of seismic control of structures. Some researchers
applied the controller to control smart base-isolated structures during earthquakes [19–22].
They are also implemented for the seismically excited structures equipped with ATMD [23–
27]. Shahi et al. [14] developed the PID controller for the structures equipped with ATMD
considering the SSI effects.
The PID controllers often represent a satisfactory performance without any improvements.
In other words, they may result in poor performance for some applications. To enhance the
performance of PID controllers, various strategies are suggested. A useful and powerful strat-
egy is the idea of the combination of fractional calculus with the conventional PID controller
that represents a new expansion of the conventional PID controller, namely FOPID con-
trollers. An appropriate setting of fractional integrator and differentiator actions can be from
the FOPID controller. The FOPID controllers have been successfully implemented in different
engineering applications. The advantages of the FOPID controller in terms of flexibility, ro-
bustness against model uncertainties, and good disturbance rejection make its implementation
for mitigation and control of vibration [28–30]. The FOPID controllers have been recently
developed in the area of seismic control of civil structures. Zamani et al. [31] investigated
the performance of the optimal FOPID controller tuned by a multi-objective cuckoo search
approach for smart base-isolated structures during both far-field and near-field ground mo-
tion. Adaptive tuning of fractional order fuzzy PID controller is developed for smart seismic
isolated structure equipped with variable friction dampers VFD. They designed a sub-level
fractional order fuzzy PID controller to reduce the isolator displacement without a significant
increase in roof acceleration and tuned online the fuzzy rule weights based on information
sensed from both the ground acceleration and the structural responses [32]. Zamani et al.
[33] proposed an adaptive fractional order fuzzy PID control approach for smart base-isolated
structures equipped with magneto-rheological dampers. A GBMO-based FOPID controller
is also proposed for tuning the control force of an ATMD embedded on the top floor of
an 11-story building [34]. Considering the time-delay effects, an optimal FOPID controller
is successfully implemented for a multi-input multi-output seismic-excited structural system
[35].
Recently, some researchers attempt to study fault detection approaches, the influence of
disturbances, modeling errors, and various uncertainties in real systems that could promote
reading across many fields, especially in the field of seismic control of structures. Zhang et
al. [36] investigated the problem of fault detection filter design for a class of inhomogeneous
higher-level Markov jump systems with uncertain transition probabilities. Xilin et al. [37] pro-
posed a novel online mode-free integral reinforcement learning algorithm is proposed to solve

10538
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

the multiplayer non-zero sum games. The proposed policy iterative algorithm can solve alge-
braic equations related to multiplayer non-zero sum games. Cheng et al. [38] studied a novel
processing method based on the matrix transformation to deal with the double summation in-
equality containing fuzzy weighting functions based on regard to the T–S fuzzy model. They
concluded that the resulting synchronization error system is mean-square exponentially stable
with a prescribed H∞ performance in the presence of actuator gain variations. Wang et al.
[39] introduced a model-based fuzzy filtering of a discrete-time semi-Markov jump nonlinear
system in which the random jumps are governed by the discrete-time semi-Markov process.
Wang et al. [40] studied the nonfragile H∞ synchronization issue for a class of discrete-time
Takagi–Sugeno fuzzy Markov jump systems. They considered the transition probabilities of
the Markov chain as piecewise time-varying and described the variation characteristics by the
persistent dwell-time switching regularity. Liu et al. [41] explored the problem of designing a
passive filter for nonlinear switched singularly perturbed systems with parameter uncertainties.
a tunnel diode electronic circuit is rendered as an example to confirm the correctness and
validity of the developed filter.
The motivation of the present study is to implement the OFOPID controller for the struc-
tures equipped with ATMD including the SSI effects. For this aim, a 15-story structure is
selected and the governing dynamic equations of the structures equipped with ATMD includ-
ing the SSI effect are extracted and an optimal design procedure is applied for tuning the
OFOPID controller. A suitable controller should reduce the structural response significantly
during earthquake excitation. The considerable displacement of the structure’s top floor is a
major concern for the designers. In addition, reducing the acceleration of floors, especially
for tall structures, should be considered to reduce its destructive effects on the structure’s
equipment. The evaluation of the seismic performance of the proposed OFOPID controller
shows this controller can remarkably reduce the maximum floor displacement at the cost of an
increase in the maximum floor acceleration in different soil conditions. Hence, for overcoming
this issue, a new modification of the controller is proposed in some way to track the amount of
the maximum floor acceleration of the structure and consider this seismic response in tuning
the demand control force of the actuator. The proposed controller, namely the OFOPID con-
troller, utilized a fuzzy block to combine the effect of roof acceleration and roof velocity as
the input of the FOPID controller. Considering four well-known real earthquakes, the seismic
performance of the proposed controllers i.e. the OFOPID and OFPD-FOPID controllers are
investigated. For this purpose, time history responses, the maximum seismic responses, and
their maximum RMS values in terms of floor displacement, acceleration, and drift responses
are compared for the different conditions of soils including dense, medium, and soft soils.
The remainder of the study is organized as follows: Mathematical model of a structure
equipped with ATMD including SSI effects is introduced in the second section. The method-
ology of the OFPD-FOPID controller is proposed in the third section. The simulation results
are discussed in Section 4. The concluding remarks are also summarized in Section 5.

2. Mathematical model

A 15-story building equipped with an ATMD palace on the top floor, including SSI ef-
fects, is selected for the numerical study. The moving of the TMD added one degree of
freedom (DOF) and both translation and rotational movement of the foundation added two
DOF to the whole system (the building with an ATMD including SSI effects). The breadth,
depth, and height of the structure are 23 m, 23 m, and 52.5 m, respectively, The story

10539
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Table 1
The parameters of soil and foundation [42–44].

Soil type Dense Medium Soft

Poisson’s ratio 0.33 0.48 0.49
Soil density (kg/m3 ) 2400 1900 1800
Shear wave velocity (m/s) 500 300 100
Shear modulus (N/m2 ) 6.00 × 108 1.71 × 108 1.80 × 107
Swaying damping (N.s/m) 109 × 1.32 108 × 6.90 108 × 2.19
Swaying stiffness (N/m) 1010 × 5.75 1010 × 1.80 109 × 1.91
Rocking damping (N.s.m) 1011 × 1.15 1010 × 7.02 1010 × 2.26
Rocking stiffness (N. m) 1.91 × 1013 7.02 × 1012 7.53 × 1011

height, story elevation relative to the ground, story mass, and story mass moment of iner-
tia are h = 3.5 m, Hi = 3.5 × i, mi = 3.456 × 105 kg, Ii = 0.146 × 108 kgm2 , respec-
tively where i = 1, . . . , 15. The stiffness coefficients of the floors are k1 = 18, 050, 000 N/m,
k j = 340, 400, 000 N/m, where j = 2, . . . , 15. Moreover, the damping coefficients of the
floors are c1 = 26, 170 N.s/m, c j =293,700 N.s/m, where j = 2, . . . , 15 [24]. The foundation
mass and foundation mass moment of inertia are designed as M0 = 0.655 × 106 kg, and
I0 = 0.22 × 108 kgm2 , respectively [24,42]. The values of the foundation’s swaying damp-
ing and stiffness (cs and ks ), as well as the foundation’s rocking damping and stiffness (cr and
kr ), are determined by soil properties including Poisson’s ratio, density, shear wave velocity,
shear modulus of the soils, and the foundation radius [42–44]. Three soil conditions: dense,
medium, and soft soil are considered in this study. The corresponding parameters of the soil
and foundation are summarized in Table 1.
Using the Lagrangian method, the mathematical model of the equation of motion of the
structure can be given by the following equation [14,18]:
M ẍ(t ) + C x˙ (t ) + Kx(t ) = −M̄ I üg (t ) + D f (t ) (1)
where M, C, and K are the mass, damping, and stiffness matrices of the whole system,
respectively, x(t ) is a vector including the floor’s displacements, TMD’s stroke, displacement,
and rotation of the foundation. Also, M̄ represents the matrix of acceleration mass for the
ground acceleration üg (t ). Moreover, I and D represent the location vectors of earthquake load
and control force of the ATMD, f (t ) is the control force of the ATMD. The above-mentioned
matrices and vectors for the studied structure are given in Appendix 1. The fundamental
frequencies of the structure are 1.67 r ad/s, 1.66 r ad/s, and 1.59 r ad/s for the dense, medium,
and soft soils, respectively.
The TMD parameters have an effective role in the seismic performance of the structure
equipped with it. Using an optimal design procedure based on the PSO algorithm, the param-
eters of the passive TMD are tuned considering the following cost function:
GOT MD = GOT
1
MD
+ GOT
2
MD
(2)
where
maxt xW ith
Roo f
T MD
(t )
GOT
1
MD
= (3)
maxt xW ithout T MD t 
Roo f ( )
maxt aW ith
Roo f
T MD
(t )
GOT
2
MD
= (4)
maxt aW ithout T MD t 
Roo f ( )

10540
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 1. Time history of the artificial ground acceleration record.

Table 2
Optimal TMD parameters for different soil types.

Soil type COT MD (N.s/m ) kOT MD (N/m )

Dense 1367.912 180639.184
Medium 3654.251 137459.365
Soft 5412.221 167459.512

in which GOT 1
MD
represents the maximum displacement of the top floor of the structure
equipped with TMD normalized by the corresponding responses in the structure without
TMD. Similarly, GOT 2
MD
refers to the maximum acceleration of the top floor of the structure
equipped with TMD normalized by the corresponding responses in the structure without
TMD. The optimal design vector is OT d
MD
= [ cOT MD kOT MD ]T for a preselected TMD mass
of 104.918 tons [23–24], which includes the optimal damping, and stiffness coefficients of the
T
TMD. The upper band vector of the design variables is selected as [ 1000 kN.s m
5000 kN
m
] .
In the optimal design procedure of the TMD parameters, the structure is subjected to an
artificial ground acceleration record simulated using the non-stationary Kanai–Tajimi model.
The artificial ground acceleration is shown in Fig. 1. The optimal design of TMD for the
studied structure located on three soil: dense, medium, and soft soils are obtained according
to Table 2.
3. Methodology of the proposed OFPD-FOPID controller

An introduction to fractional calculus is given in the first subsection. Then, how the im-
plementation of the OFOPID controller for the studied structure for seismic applications is
described. The motivation and methodology of the proposed OFPD-FOPID controller to over-
come some issues excited in the FOPID controller are given at the end.

3.1. An introduction to fractional calculus

Fractional calculus is a non-integer order fundamental operator generalization of integration

and differentiation a Dtα as follows:
⎧ dα
⎨ dt α α>0
α
a Dt = 1 α =0 (5)
⎩ t α
a ( dτ ) α < 0
10541
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

The operation’s boundaries are a and t, and its order is α(α ∈ R ). For fractional inte-
gration and differentiation, several definitions have been suggested, including the Riemann–
Liouville (RL), Grünwald–Letnikov (GL), and Caputo. Many control applications in the lit-
erature employ Caputo’s fractional differentiation formula to derive fractional-order transfer
function models from fractional-order ordinary differential equations with zero initial condi-
tions [45,46]. According to this formula, the α order derivative of a function f (t ) to time is
given by:
 t
1 Dm f (t )
Dα f (t ) = dτ, α ∈ R+ , m ∈ Z + , m − 1 ≤ α ≤ m (6)
(m − α) 0 (t − τ )α+1−m
The Laplace transform of Eq. (4) can be written as:
 ∞
  m−1
L a Dtα = e−st Dα f (t )dt = sα F (s ) − sα−k−1 Dk f (0 ) (7)
0 k=0

where L{·} is the Laplace transform of f (t ) and (·) is the Gamma function.
Numerical approximations of fractional integrals and derivatives are the most commonly
used operators in realistic implementations and simulations. The concept is to use integer-
order transfer functions to approximate the behavior of the fractional-order Laplace operator.
Oustaloup’s approach is a well-known and accurate frequency-domain method for making this
estimate [47–48]. This approximation employs a recursive distribution with N poles and N
zeroes. The Oustaloup approximation is described in the frequency interval [ωl , ωh ] for N
corner frequencies [48].
N 1+ s
ωz,n
sα ≈ k ,α > 0 (8)
n=1
1+ s
ω p,n

in which the gain k is used to adjust both sides of Eq. (6) with unit gain at 1 rad/s. Oustaloup
[47–50] was the first to utilize fractional-order controllers (FOC) to control dynamic systems,
developing the so-called CRONE (French abbreviation for "Commande Robuste d’Ordre Non-
Entier") controller. He demonstrates how this controller outperforms the conventional PID
controller.

3.2. Implementation of the OFOPID controller

Podlubny firstly [51] proposed the FOPID or PI λ Dμ controller, which is a generalization

of the PID controller. Fractional calculus and fractional controller design are performed in
either the time domain or the frequency domain. The FOPID controller’s time-domain calcu-
lation is extremely complex. As a result, the FOPID controller’s frequency domain form is
widely employed in engineering applications. The time-domain output of the controller and
the transfer function of the FOPID controller are given by:
u (t ) = KP e(t ) + KI D−λ e(t ) + KD Dμ e(t ) (9)

KI
KF OPID (s ) = KP + + KD sμ (10)
sλ
where KP , KI and KD represent proportional, integral, and derivative gains. In addition, λ and
μ are the orders of integral and derivative operators, respectively. The duration of the control

10542
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

procedure is specified by the parameter t. In comparison to the typical PID controller, the
FOPID controller has two additional parameters. As a result, tuning the FOPID controller
can be significantly more complex. The integrative and derivative orders of the classic PID
controller are shifted from one point to the λ − μ plane by the FOPID controller. To put
it another way, PD, PI, and PID controllers only cover the corner, whereas fractional PID
controllers cover the entire region. The extension gives you more control over the design of
your controls. The fractional-order PID controller has five parameters to regulate, whereas
the PID controller has three. Furthermore, the FOPID controller improves the typical PID
controller in terms of load disturbance rejection and setpoint regulation. Furthermore, it is
less sensitive to model parameter changes. These benefits of the FOPID controller are crucial
in vibration control systems, and they also mean that the FOPID controller can be used to
mitigate vibration in seismically excited structures.
A FOPID controller’s efficacy is related by the proportional, integral, and derivative coef-
ficients, as well as the order of the integral and derivative operator. The design of a FOPID
controller could be viewed as a search problem. The vector OF d
OPID
= [KP , KI , KD , λ, μ]T is
the design variable vector. An optimization problem can be defined concerning the minimiz-
ing of performance indices of the seismic-excited structure. However, determining optimal
control parameters is difficult due to uncertainty in the external excitation and structural sys-
tems. The optimization methods based on evolutionary algorithms can be used to solve this
challenge. The integer-order approximation transfer function of the FOPID controller must be
calculated using the Oustaloup technique for implementation of the FOPID controller toward
seismic control of the structure equipped with an ATMD system. The parameters utilized in
the Oustaloup approximation are ωl = 0.01,ωh = 100 and N = 6.
A suitable controller should reduce the structural response significantly during earthquake
excitation. The considerable displacement of the structure’s top floor is a major concern for
the designers. In addition, reducing the acceleration of floors, especially for tall structures,
should be considered to reduce its destructive effects on the structure’s equipment. Two goals
should be delivered with considering a saturation for the control force from the practical point
of view. For practical engineering applications, the maximum control force of the ATMD is
limited to 0.05W in which W is the total weight of the structure [14,18]. As a result, a cost
function including the mentioned goals is addressed in the design procedure of the OFOPID
controller as follows:
GOF OPID = GOF
1
OPID
+ GOF
2
OPID
(11)
where
maxt xControl
Roo f
l ed
(t )
GOF
1
OPID
= (12)
maxt xU ncontrol l ed t 
Roo f ( )
maxt aControl
Roo f
l ed
(t )
GOF
2
OPID
= (13)
maxt aURoo
ncontrol l ed t 
f ( )
here, GOF1
OPID
is the maximum displacement of the top story in the controlled struc-
ture (equipped with ATMD) subjected to the artificial ground acceleration normalized
to its response in the uncontrolled structure (equipped with passive TMD). Additionally,
GOF
2
OPID
shows the maximum roof acceleration in the uncontrolled structure normalized to
its corresponding value in the uncontrolled structure. The lower and upper bounds of the
OFOPID controller parameters are 0 ≤ KP ≤ 109 , 0 ≤ KI ≤ 109 , 0 ≤ KD ≤ 109 , 0 ≤ λ ≤ 2

10543
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 2. The block diagram of the proposed OFOPID controller.

Table 3
Optimal parameters of the OFOPID controller for different soil types.

Soil type KP Kt KD λ μ
Dense 36495337.852 193957799.879 48743940.419 1.726 0.298
Medium 178466243.522 457852227.505 42300243.687 1.738 0.541
Soft 4872029.155 42498134.051 804307.626 1.008 0.344

Table 4
Optimal parameters of the optimal PID controller for different soil types.

Soil type KP Kt KD
Dense 2432595.566 44622949.228 2501952.297
Medium 6517578.381 53082552.729 7475808.795
Soft 1442913.205 53291607.743 398613.864

and 0 ≤ μ ≤ 2. The maximum control force of the ATMD is considered a satiation block in
MatLab/Simulink software which represents a nonlinear control design of the FOPID con-
troller. According to [34], the best input for the FOPID controller is the top floor velocity
to implement the PID and FOPID controller. The block diagram of the proposed OFOPID
controller is shown in Fig. 2. In the proposed block diagram, the measured top-floor velocity
is used as the input of the proposed OFOPID controller. The output of the proposed OFOPID
applies as an active control force to the tuned mass damper located on this floor.
According to different soil types, the proposed OFOPID controllers are designed for the
structure using the PSO algorithm. Table 3 indicates the optimal values of the proposed
OFOPID controllers for different soil types.
For comparison purposes, an optimal PID controller is also designed. The design space of
KP , KI , and KD and the optimization cost function are intended to be similar to what they
are intended for the OFOPID design procedure. The optimal parameters of the optimal PID
controllers for different soil types are shown in Table 4.
One of the most widely used controllers in the area of seismic control of structure is the
fuzzy PID (FPID) controller. The structure of this controller is illustrated in Fig. 3. In the

10544
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 3. The block diagram of the FPID controller and its MFs for the inputs and output variables.

Table 5
The rule base of the FPID controllers.
dvroo f
vroo f / dt NB NM NS Z PS PM PB
NB NB NB NB NB NM NS Z
NM NB NB NB NM NS Z PS
NS NB NB NM NS Z PS PM
Z NB NM NS Z PS PM PB
PS NM NS Z PS PM PB PB
PM NS Z PS PM PB PB PB
PB Z PS PM PB PB PB PB

designed FPID controller, seven membership functions as shown in Fig. 3 are considered for
the inputs and output. The table of rules for this block is depicted in Table 5. The vector
Fd PID = [Ke , Kd , α, β]T is the design variable vector. The lower and upper bounds of the
FPID controller parameters are 0 ≤ Ke ≤ 1, 0 ≤ Kd ≤ 1, 0 ≤ α ≤ 109 , and 0 ≤ β ≤ 109 . The
cost function in the optimal design of the controller is considered the same as the OFOPID
design procedure. The optimal parameters of the optimal FPID controllers for different soil
types are inserted in Table 6.

10545
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Table 6
Optimal parameters of the FPID controller for different soil types.

Soil type Ke Kd α β
Dense 0.631 0.302 217232735.961 50206258.633
Medium 0.619 0.267 434959616.136 43569251.981
Soft 0.556 0.219 45048022.092 764092.244

Fig. 4. The block diagram of the proposed OFPD-FOPID controller.

3.3. Implementation of the proposed OFPD-FOPID controller

According to the results obtained from the OFOPID controller implemented for the struc-
ture subjected to the artificial earthquake, a remarkable reduction is given for floor displace-
ment. However, it is obtained at the cost of an increase of 56%, 86%, and 79% in the
maximum floor acceleration in different soil conditions. Hence, it is far from a suitable con-
troller especially for structures with sensitive equipment to vibration. To overcome this issue,
a new modification OFOPID controller is suggested in this study that tracks the amount of
the maximum floor acceleration of the structure and takes into account it in determining
the amount of the control force commanded to the actuator. For this purpose, the proposed
controller, namely the OFOPID controller, uses a fuzzy block to combine the effect of roof
acceleration and roof velocity as the input of the FOPID controller as illustrated in Fig. 4.
In the proposed block diagram, the measured roof velocity and roof acceleration apply to
a fuzzy logic controller (FLC). The FLC block is a Mamdani fuzzy inference system. The
product implication, three triangular and two trapezoidal membership functions, and the cen-
troid defuzzification method are the structural configuration of the FLC block in the proposed
OFPD-FOPID.
Considering the studied structure during the artificial ground acceleration, the pro-
posed OFPD-FOPID controller is implemented for the studied structures. As can be seen
from Fig. 4, the design parameters vector for this control strategy is OF d
PD−F OPID
=
[K1 , K2 , KP , KI , KD , λ, μ] . The intervals considered for designing the controller parame-
T

ters are 0 ≤ K1 ≤ 1, 0 ≤ K2 ≤ 1, 0 ≤ KP ≤ 109 , 0 ≤ KI ≤ 109 , 0 ≤ KD ≤ 109 , 0 ≤ λ ≤ 2 and

0 ≤ μ ≤ 2. According to Eqs. (9) to (11), the same objective function is also considered to

10546
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 5. The used MFs for the input and output variables.

Table 7
The rule base of the proposed controllers.

vroo f /aroo f NB NM Z PM PB
NB NB NB NB NM Z
NM NB NB NM Z PM
Z NB NM Z PM PB
PM NM Z PM PB PB
PB Z PM PB PB PB

Table 8
Optimal parameters of the OFPD-FOPID controller for different soil types.

Soil type K1 K2 KP KI KD λ μ
Dense 0.697 0.267 159155922.289 304972742.198 537569.467 1.865 0.817
Medium 0.599 0.204 80063525.647 351138277.916 2610926.361 1.890 0.681
Soft 0.531 0.216 552403.431 6064686.144 617623.922 1.100 0.041

design the OFPD-FOPID controller. The fuzzy block used in the proposed control strategy
has five membership functions as shown in Fig. 5 for the inputs and output. The table of rules
for this block is depicted in Table 7. The optimal parameters of the OFPD-FOPID controller
for different soil types are obtained using the PSO algorithm. Table 8 represents the optimal
parameters of the OFPD-FOPID controller for different soil types.
The corresponding cost functions given for the PID, OFOPID, and OFPD-FOPID con-
trollers in Table 9 indicate the performance superiority of the proposed OFPD-FOPID con-
troller over the OFOPID and the PID controller in simultaneous reduction of the maximum
floor displacement and acceleration. As shown in Fig. 6, a comparison of the time responses
of the uncontrolled structure and controlled ones using the PID, OFOPID, and OFPD-FOPID

10547
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 6. Time responses of the structure located on (a) dense, (b) medium, and (c) soft soils during the artificial
ground acceleration.

10548
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 6. Continued

Table 9
The cost function values of the structure during the artificial earthquake.

PID FPID OFOPID OFPD-FOPID

Soil types G1 G2 G G1 G2 G G1 G2 G G1 G2 G
Soft soil 0.39 1.68 2.07 0.34 1.23 1.57 0.21 1.56 1.77 0.17 0.89 1.06
Medium soil 0.39 2.33 2.72 0.35 1.75 2.10 0.13 1.86 1.99 0.11 0.90 1.01
Dense soil 0.32 1.93 2.25 0.28 1.81 2.09 0.17 1.79 1.96 0.13 0.91 1.04

controllers during the artificial earthquake in different soil conditions has also confirmed this
result.

4. Results and discussion

The seismic performance of the proposed controllers i.e. the OFOPID and OFPD-FOPID
controllers as well as the optimal PID and FPID controllers are evaluated in different real
well-known earthquakes. Four ground acceleration data extracted from the El-Centro, Kobe,
Northridge, and Hachinohe earthquakes are selected for this purpose. The maximum responses
of the structures in both uncontrolled and controlled with the proposed controllers are com-
pared for the different conditions of soils including dense, medium, and soft soils, respectively
in Figs. 7, 8, and 9.
Fig. 7 illustrates a bar graph of the maximum floor displacement of the structure subjected
to the mentioned four earthquakes in both uncontrolled and controlled cases with the PID,

10549
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 7. Maximum floor displacement of the uncontrolled structure and controlled ones using the proposed controllers
in different soil types.

10550
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 8. Maximum floor acceleration of the uncontrolled structure and controlled ones using the proposed controllers
in different soil types.

10551
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 9. The maximum inter-story drift of the uncontrolled structure and controlled ones using the proposed controllers
in different soil types.

10552
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

FPID, and the proposed OFOPID and OFPD-FOPID controllers in three conditions of dense,
medium, and soft soils. During the Kobe earthquake as a near-field earthquake, a reduction of
34%, 34%, 27%, and 43% in the maximum floor displacement results from the PID, FPID,
OFOPID, and OFPD-FOPID controllers over the uncontrolled case in dense soil. During
the Hachinohe earthquake, the PID, FPID, OFOPID, and OFPD-FOPID controllers give a
reduction of 52%, 58%, 59%, and 66%, respectively. It is also concluded the PID, FPID,
OFOPID, and OFPD-FOPID controllers can reduce the maximum floor displacement of the
structure located on the medium soil by 10%, 18%, 34%, and 46% during the Kobe earthquake
and 29%, 45%, 66% and 73% during the Hachinohe earthquake. Considering the soft soil,
these reductions are 31%, 36%, 44%, and 51% for the Kobe earthquake and 11%, 18%, 30%,
and 43% for the Hachinohe earthquake. Consequently, the proposed OFPD-FOPID controller
represents a better performance than the PID, FPID, and OFOPID controllers in the reduction
of the maximum floor displacement. The results also show an increment in seismic response
both uncontrolled and controlled by increasing the soil softness.
A bar graph of the maximum floor acceleration of the structure is displayed in Fig. 8 in
different earthquakes and three soil types. The superiority of the proposed OFPD-FOPID con-
troller over the PID and OFOPID controllers in this figure is evident. Although the OFOPID
controller was able to reduce the maximum floor displacement; it is achieved at the cost of
an increment in the maximum floor acceleration of the structure. Considering the structure
subjected to the Kobe earthquake, for the FOPID controller an increase of 41%, 46%, and
25%, in the PID controller an increase of 42%, 54%, and 16% have resulted in maximum
floor acceleration tuned for the dense, medium, and soft soils, respectively. Similarly, the PID
controller makes an increase of 41%, 79%, and 63% during the Hachinohe earthquake for
the dense, medium, and soft soils. Furthermore, the FPID controller causes increases of 41%,
79%, and 63% and the OFOPID controller makes an increase of 59%, 73%, and 30%, respec-
tively, while the OFPD-FOPID controller slightly decreases the maximum floor acceleration
in different soil types.
Fig. 9 demonstrates a bar graph of the maximum inter-story drift of the uncontrolled
and controlled structure in three types of soils: dense, medium, and soft soils. It can be
seen that all controllers are significantly able to reduce the maximum inter-story drift. The
results show that the OFPD-FOPID controller performs better than another controller in the
reduction of the maximum inter-story drift of the structure. For example, in dense soil, during
the Kobe earthquake, a reduction of 5%, 22%, 31%, and 39% are obtained from the PID,
FPID, OFOPID, and OFPD-FOPID controllers than the uncontrolled structure. The results
given for the structure located on the medium soil show that the maximum inter-story drifts
of the structure subjected to the Kobe earthquake are reduced by about 10%, 16%, 31%,
and 43% using the PID and OFOPID and OFPD-FOPID controllers, while these reductions
are obtained in soft soil as 29%, 33%, 40% and 46% for the Kobe earthquake respectively.
Similarly, these reductions are achieved by 32%, 44%, 50%, and 54% for the Hachinohe
earthquake, respectively. These reductions are 20%, 28%, 44%, and 55% when the structure
subjected to the Hachinohe earthquake is located on medium soil. It is also found that the
maximum inter-story drifts of the structure located on the soft soil experience a reduction of
13%, 14%, 21%, and 28% during the Hachinohe earthquake using the PID, FPID, OFOPID,
and OFPD-FOPID controllers, respectively.
The RMS values of the seismic responses of the structure located on the dense, medium,
and soft soils are summarized in Table 10 during different earthquake excitations. The max-
imum RMS values of the top floor displacement, acceleration, and drifts are reported in the

10553
 S. Etedali, A.-A. Zamani, M. Akbari et al.
Table 10
Maximum RMS values of the seismic responses of the structure located on different soil types.

Uncontrol PID FPID OFOPID OFPD-FOPID

Drift Disp. Acc. Drift Disp. Acc. Drift Disp. Acc. Drift Disp. Acc. Drift Disp. Acc.
(m) (m) (m/s2 ) (m) (m) (m/s2 ) (m) (m) (m/s2 ) (m) (m) (m/s2 ) (m) (m) (m/s2 )
Dense El Centro 0.050 0.220 3.327 0.047 0.189 5.188 0.035 0.139 5.071 0.028 0.105 5.489 0.026 0.010 3.568
Kobe 0.092 0.386 8.879 0.091 0.314 11.115 0.074 0.270 10.558 0.066 0.255 11.113 0.062 0.232 6.890
10554

Northridge 0.140 0.666 9.787 0.136 0.607 10.361 0.122 0.550 10.003 0.122 0.551 10.699 0.112 0.513 7.596

Journal of the Franklin Institute 360 (2023) 10536–10563

Hachinohe 0.062 0.283 2.874 0.043 0.167 3.478 0.035 0.137 3.602 0.032 0.122 4.106 0.030 0.112 3.080
Medium El Centro 0.051 0.219 3.321 0.049 0.195 5.801 0.041 0.154 6.001 0.029 0.100 5.631 0.023 0.082 3.266
Kobe 0.092 0.410 8.804 0.081 0.389 12.418 0.077 0.338 11.198 0.066 0.256 11.388 0.053 0.212 8.655
Northridge 0.139 0.669 9.792 0.128 0.545 11.016 0.132 0.583 11.441 0.125 0.566 10.778 0.101 0.458 9.161
Hachinohe 0.055 0.262 2.875 0.045 0.192 4.211 0.040 0.161 4.143 0.032 0.114 4.443 0.025 0.091 2.843
Soft El Centro 0.047 0.204 3.332 0.046 0.181 4.155 0.044 0.165 4.235 0.040 0.141 4.068 0.034 0.114 3.173
Kobe 0.096 0.450 9.077 0.069 0.344 10.264 0.066 0.322 10.673 0.059 0.281 10.460 0.048 0.222 8.891
Northridge 0.118 0.596 9.407 0.117 0.586 10.899 0.119 0.586 10.936 0.115 0.552 10.336 0.093 0.441 9.199
Hachinohe 0.051 0.242 2.867 0.049 0.226 3.731 0.046 0.201 3.491 0.041 0.164 3.047 0.032 0.129 2.743
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 10. Time responses of the structure located on (a) dense, (b) medium and (c) soft soils during the El-Centro
earthquake.

10555
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 10. Continued

figure. The simulation results show that the maximum RMS values of the top floor displace-
ment and drift are remarkably decreased using the PID, FPID, and OFOPID controller, while
a remarkable increase is obtained for the maximum RMS values of the top floor acceleration
for different soil types and ground accelerations. Nevertheless, the proposed OFPD-FOPID
controller has been able to overcome the problem of increasing the acceleration of the floor in
another controller. In other words, in addition to a very noticeable reduction in the maximum
RMS values of the top floor displacement and drifts, it is slightly able to reduce the maximum
RMS values of the top floor displacement during the most studied earthquakes.
Time history responses of the structure during the El-Centro and Northridge are compared
in Figs. 10 and 11, respectively. In El-Centro as a far-field earthquake, the maximum top
floor displacements are reduced by 60%, 58%, and 52% for the OFOPID controller and 30%,
4%, and 15% for the PID controller in dense, medium, and soft soils, respectively, while the
OFOPID gives an increment of 38%, 42%, and 1% for the OFOPID controller and 45%,
47%, and 4% for the PID controller in the maximum top floor accelerations, respectively.
Similarly, the proposed OFPD-FOPID controller gives a reduction of 67%, 66%, and 61% in
the maximum top floor displacements, while a reduction is also given in the maximum top
floor acceleration. During the Northridge as a near-filed earthquake, a reduction of 18%, 20%,
and 5% for the OFOPID and a reduction of 7%, 26%, and 0.1% for the PID controller in the
maximum top floor displacements is observed in dense, medium, and soft soils, respectively,
while the OFOPID gives an increment of 1%, 6%, and 18% and the PID controller gives
an increment of 7%, 35%, and 22% in the maximum top floor accelerations, respectively.
Similarly, the proposed OFPD-FOPID controller results in a reduction of 34%, 35%, and

10556
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 11. Time responses of the structure located on (a) dense, (b) medium, and (c) soft soils during the Northridge
earthquake.

10557
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

Fig. 11. Continued

25% in the maximum top floor displacements in dense, medium, and soft soils, respectively,
while a reduction is also given in the maximum top floor acceleration. Consequently, both
controllers are capable of reducing the top floor displacement. It can be seen that the PID
and OFOPID controllers result in an increment in top floor acceleration, the OFPD-FOPID
controller while maintaining its significant effect in reducing the top floor displacement, has
also slightly reduced the top floor acceleration.
One velocity sensor is needed to implement the proposed PID and OFOPID control strate-
gies. For the proposed OFPD-FOPID, one velocity, and one acceleration sensor are required.

5. Conclusions

In this paper, an OFOPID controller was first implemented for seismic control of a 15-
story structure equipped with an ATMD including SSI effects. The simulation results showed it
could remarkably reduce the maximum floor displacement over the conventional PID controller
at the cost of an increase in the maximum floor acceleration of the structure located on differ-
ent soil conditions. To overcome the issue of increasing floor acceleration, a new framework
of optimal FOPID controller series with fuzzy PD controller, namely OFPD-FOPID controller,
was developed. Considering four real earthquakes and three soil types, the OFOPID controller
gives an average reduction of 41%, 45%, and 33% in the maximum floor displacement of
the structure located on the dense, medium, and soft soils, while the proposed OFPD-FOPID
controller represents an average reduction of 52%, 55%, and 45% in the maximum floor dis-

10558
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

placement. Similarly, an average reduction of 36%, 33%, and 20% resulted in the OFOPID
controller in the maximum inter-story drift, while the proposed OFPD-FOPID controller re-
duced this seismic response by 42%, 44%, and 28%. Hence, the OFPD-FOPID controller has
superiority over the OFOPID controller in the reduction of maximum floor displacement and
inter-story drift. While the OFOPID controller gave a poor performance in mitigation of the
maximum floor acceleration and increased it by 45%, 52%, and, 24% in dense, medium, and
soft soils, the proposed OFPD-FOPID controller represented a slight reduction in this seismic
response. Furthermore, the OFOPID controller gave a remarkable increase in the maximum
RMS values of the top floor acceleration for different soil types and ground accelerations.
Nevertheless, the proposed OFPD-FOPID controller had been able to overcome the problem of
increasing the acceleration of the roof floor. Although a performance decline was often found
for the OFOPID controller in near-field earthquakes, the proposed OFPD-FOPID controller
was able to maintain its appropriate performance in these earthquakes.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.

Appendix A

The mentioned matrices and vectors in Eq. (1) for the studied structure are given as follows:

⎡ ⎤
[M ]15 × 15 {0}15 × 1 {M }15 × 1 {MH }15 × 1
⎢ ⎥
⎢ mT MD mT MD mT MD H15 ⎥
M=⎢
⎢
⎥
⎥
⎣ M0 +  15
j=1 (m j ) + mT MD i=1
15
(mi Hi ) + mT MD H15 ⎦
 
Sym. I0 + i=1
15
Ii +mi Hi2 +mT MD H15
2

(A1)

⎡   ⎤
0(15−1) × 1
⎢[C ]15 × 15 {0}15 × 1 {0}15 × 1 ⎥
⎢ −cT MD ⎥
C=⎢ ⎢ cT MD 0 0 ⎥
⎥ (A2)
⎣ cs 0 ⎦
Sym. cr
⎡   ⎤
0(15−1) × 1
⎢[K ]15 × 15 {0}15 × 1 {0}15 × 1 ⎥
⎢ −kT MD ⎥
K=⎢ ⎢ kT MD 0 0 ⎥
⎥ (A3)
⎣ ks 0 ⎦
Sym. kr
⎡ ⎤
[M ]15 × 15 {0}15 × 1 {0}15 × 1 {0}15 × 1
⎢ mT MD 0 0 ⎥
M̄ = ⎢
⎣
15 ⎥ (A4)
⎦
M0 + i=1 (mi ) + mT MD 0
N
Sym. i=1 (mi Hi ) + mT MD H15

10559
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

⎧ ⎫
⎪
⎪ x1 (t ) ⎪
⎪
⎪
⎪ ⎪
⎪
⎪ x2 (t ) ⎪
⎪
⎪
⎪
⎪ ⎪
⎪
⎪
⎨
.. ⎪
⎬
.
x(t ) = x15 (t ) ⎪ (A5)
⎪
⎪ ⎪
⎪
⎪ x ⎪
T MD (t )⎪
⎪
⎪ ⎪
⎪
⎪
⎪ x0 (t ) ⎪
⎪
⎪
⎩ ⎪
⎭
θ0 (t )
⎧ ⎫
⎪
⎪ 0 ⎪
⎪
⎪
⎪ ⎪
⎪
⎪
⎪ 0 ⎪
⎪
⎪
⎪ ⎪
⎪
⎪
⎨
.. ⎪
⎬
.
D= −1⎪ (A6)
⎪
⎪ ⎪
⎪
⎪ 1⎪
⎪
⎪
⎪ ⎪
⎪
⎪
⎪ 0 ⎪
⎪
⎪
⎩ ⎪ ⎭
0

where
⎡ ⎤ ⎧ ⎫
m1 0 0 0 0 ⎪
⎪ m1 ⎪⎪
⎢ ⎪ ⎪
⎢ m2 0 0 0 ⎥ ⎥
⎪
⎪
⎨ 2⎪
m ⎪
⎬
⎢ .. ⎥ .
[M ]15 × 15 =⎢ . {0 } {0 } ⎥ ; {M }15 × 1 = .. ; {MH }15 × 1
⎢ ⎥ ⎪
⎪ ⎪
⎪
⎣ m14 0 ⎦ ⎪
⎪ m14 ⎪
⎪
⎪
⎩ ⎪ ⎭
Sym. m15 m15
⎧ ⎫
⎪
⎪ m1 H1 ⎪ ⎪
⎪
⎪ ⎪
⎪
⎪
⎨ m2 H2 ⎪ ⎬
= .. (A7)
⎪ . ⎪
⎪
⎪ ⎪
⎪m14 H14 ⎪
⎪ ⎪
⎪
⎩ ⎭
m15 H15

⎡ ⎤
c1 + c2 −c2 0 0 0
⎢ .. .. ⎥
⎢ c2 + c3 . . 0 ⎥
⎢ ⎥
[C ]15×15 =⎢
⎢ .. .. ⎥
⎥ (A8)
⎢ . −c14 . ⎥
⎣ c14 + c15 −c15 ⎦
Sym. c15 + cT MD
⎡ ⎤
k1 + k2 −k2 0 0 0
⎢ .. .. ⎥
⎢ k2 + k3 . . 0 ⎥
⎢ ⎥
[K ]15 × 15 =⎢
⎢ .. .. ⎥
⎥ (A9)
⎢ . −k14 . ⎥
⎣ k14 + k15 −k15 ⎦
Sym. k15 + kT MD

10560
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

References

[1] A. Kaveh, S. Mohammadi, O.K. Hosseini, A. Keyhani, V.R. Kalatjari, Optimum parameters of tuned mass
dampers for seismic applications using charged system search, Iranian J. Sci. Technol. Trans. Civil Eng. 39
(C1) (2015) 21.
[2] S. Etedali, H. Rakhshani, Optimum design of tuned mass dampers using multi-objective cuckoo search for
buildings under seismic excitations, Alexandria Eng. J. 57 (4) (2018) 3205–3218.
[3] R. Kamgar, P. Samea, M. Khatibinia, Optimizing parameters of tuned mass damper subjected to critical earth-
quake, Struct. Design Tall Special Build. 27 (7) (2018) e1460.
[4] B. Keshtegar, S. Etedali, Nonlinear mathematical modeling and optimum design of tuned mass dampers using
adaptive dynamic harmony search algorithm, Struct. Control Health Monitor. 25 (7) (2018) e2163.
[5] A. Farshidianfar, S. Soheili, Ant colony optimization of tuned mass dampers for earthquake oscillations of
high-rise structures including soil–structure interaction, Soil Dyn. Earthquake Eng. 51 (2013) 14–22.
[6] M. Khatibinia, H. Gholami, S.F. Labbafi, Multi–objective optimization of tuned mass dampers considering
soil–structure interaction, Int. J. Optimiz. Civil Eng. 6 (4) (2016) 595–610.
[7] S. Etedali, M. Seifi, M. Akbari, A numerical study on optimal FTMD parameters considering soil-structure
interaction effects, Geomech. Eng. 16 (5) (2018) 527–538.
[8] S. Etedali, M. Akbari, M. Seifi, MOCS-based optimum design of TMD and FTMD for tall buildings under
near-field earthquakes including SSI effects, Soil Dyn. Earthquake Eng. 119 (2019) 36–50.
[9] Z. Zhao, R. Zhang, Z. Lu, A particle inerter system for structural seismic response mitigation, J. Franklin Inst.
356 (14) (2019) 7669–7688.
[10] D. De Domenico, P. Deastra, G. Ricciardi, N.D. Sims, D.J. Wagg, Novel fluid inerter based tuned mass dampers
for optimised structural control of base-isolated buildings, J. Franklin Inst. 356 (14) (2019) 7626–7649.
[11] W.M. Kuhnert, P.J.P. Gonçalves, D.F. Ledezma-Ramirez, M.J. Brennan, Inerter-like devices used for vibration
isolation: a historical perspective, J. Franklin Inst. 358 (1) (2021) 1070–1086.
[12] A.A. Taflanidis, A. Giaralis, D. Patsialis, Multi-objective optimal design of inerter-based vibration absorbers for
earthquake protection of multi-storey building structures, J. Franklin Inst. 356 (14) (2019) 7754–7784.
[13] C. Liu, L. Chen, H.P. Lee, Y. Yang, X. Zhang, A review of the inerter and inerter-based vibration isolation:
theory, devices, and applications, J. Franklin Inst. 359 (14) (2022) 7677–7707.
[14] M. Shahi, M.R. Sohrabi, S. Etedali, Seismic control of high-rise buildings equipped with ATMD including
soil-structure interaction effects, J. Earthquake Tsunami 12 (03) (2018) 1850010.
[15] H. Gao, J. Hu, Y. Qi, C. Sun, Adaptive vibration control of a flexible structure based on hybrid learning
controlled active mass damping, J. Franklin Inst. 359 (12) (2022) 5935–5959.
[16] E. Nazarimofrad, S.M. Zahrai, Fuzzy control of asymmetric plan buildings with active tuned mass damper
considering soil-structure interaction, Soil Dyn. Earthquake Eng. 115 (2018) 838–852.
[17] S. Golnargesi, H. Shariatmadar, H.M. Razavi, Seismic control of buildings with active tuned mass damper
through interval type-2 fuzzy logic controller including soil–structure interaction, Asian J. Civil Eng. 19 (2)
(2018) 177–188.
[18] S. Etedali, M. Shahi, A control scheme for AMD in the presence of time-delays and SSI effects for tall buildings,
Struct. Eng. Mech., Int. J. 79 (2) (2021) 267–278.
[19] S. Etedali, M.R. Sohrabi, S. Tavakoli, Optimal PD/PID control of smart base isolated buildings equipped with
piezoelectric friction dampers, Earthquake Eng. Eng. Vibrat. 12 (1) (2013) 39–54.
[20] S. Etedali, S. Tavakoli, M.R. Sohrabi, Design of a decoupled PID controller via MOCS for seismic control of
smart structures, Earthquakes Struct. 10 (5) (2016) 1067–1087.
[21] A.A. Zamani, S. Tavakoli, S. Etedali, Control of piezoelectric friction dampers in smart base-isolated structures
using self-tuning and adaptive fuzzy proportional–derivative controllers, J. Intell. Mater. Syst. Struct. 28 (10)
(2017) 1287–1302.
[22] J.P. Zand, J. Sabouri, J. Katebi, M. Nouri, A new time-domain robust anti-windup PID control scheme for
vibration suppression of building structure, Eng. Struct. 244 (2021) 112819.
[23] R. Guclu, H. Yazici, Vibration control of a structure with ATMD against earthquake using fuzzy logic controllers,
J. Sound Vib. 318 (1–2) (2008) 36–49.
[24] R. Guclu, H. Yazici, Seismic-vibration mitigation of a nonlinear structural system with an ATMD through a
fuzzy PID controller, Nonlinear Dyn. 58 (3) (2009) 553–564.
[25] S. Etedali, S. Tavakoli, PD/PID controller design for seismic control of high-rise buildings using multi-objective
optimization: a comparative study with LQR controller, J. Earthquake Tsunami 11 (03) (2017) 1750009.

10561
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

[26] S. Etedali, A new modified independent modal space control approach toward control of seismic-excited struc-
tures, Bull. Earthquake Eng. 15 (10) (2017) 4215–4243.
[27] A.H. Heidari, S. Etedali, M.R. Javaheri-Tafti, A hybrid LQR-PID control design for seismic control of buildings
equipped with ATMD, Front. Struct. Civil Eng. 12 (1) (2018) 44–57.
[28] S. Gad, H. Metered, A. Bassuiny, A.M. Abdel Ghany, Multi-objective genetic algorithm fractional-order PID con-
troller for semi-active magnetorheologically damped seat suspension, J. Vib. Control 23 (8) (2017) 1248–1266.
[29] L. Marinangeli, F. Alijani, S.H. HosseinNia, Fractional-order positive position feedback compensator for active
vibration control of a smart composite plate, J. Sound Vib. 412 (2018) 1–16.
[30] X. Zhang, T. Sui, H. Zhang, Y. Zhang, L. Liu, S. Zhang, An active vibration control method for typical
piping system of nuclear power plant, 2021 IEEE 10th Data Driven Control and Learning Systems Conference
(DDCLS), IEEE, 2021, pp. 385–390.
[31] A.A. Zamani, S. Tavakoli, S. Etedali, Fractional order PID control design for semi-active control of smart
base-isolated structures: a multi-objective cuckoo search approach, ISA Trans. 67 (2017) 222–232.
[32] A.A. Zamani, S. Tavakoli, S. Etedali, J. Sadeghi, Online tuning of fractional order fuzzy PID controller in smart
seismic isolated structures, Bull. Earthquake Eng. 16 (7) (2018) 3153–3170.
[33] A.A. Zamani, S. Tavakoli, S. Etedali, J. Sadeghi, Adaptive fractional order fuzzy proportional–integral–derivative
control of smart base-isolated structures equipped with magnetorheological dampers, J. Intell. Mater. Syst. Struct.
29 (5) (2018) 830–844.
[34] S. Etedali, A.A. Zamani, S. Tavakoli, A GBMO-based PIλDμ controller for vibration mitigation of seismic-ex-
cited structures, Autom. Constr. 87 (2018) 1–12.
[35] A.A. Zamani, S. Etedali, Optimal fractional-order PID control design for time-delayed multi-input multi-output
seismic-excited structural system, J. Vib. Control (2021) 10775463211053188.
[36] X. Zhang, H. Wang, V. Stojanovic, P. Cheng, S. He, X. Luan, F. Liu, Asynchronous fault detection for interval
type-2 fuzzy nonhomogeneous higher-level Markov jump systems with uncertain transition probabilities, IEEE
Trans. Fuzzy Syst. (2021).
[37] X. Xin, Y. Tu, V. Stojanovic, H. Wang, K. Shi, S. He, T. Pan, Online reinforcement learning multiplayer non-zero
sum games of continuous-time Markov jump linear systems, Appl. Math. Comput. 412 (2022) 126537.
[38] P. Cheng, H. Wang, V. Stojanovic, S. He, K. Shi, X. Luan, . . . C. Sun, Asynchronous fault detection observer
for 2-D Markov jump systems, IEEE Trans. Cybern. (2021).
[39] J. Wang, Y. Zhang, L. Su, J.H. Park, H. Shen, Model-based fuzzy filtering for discrete-time Semi-Markov jump
nonlinear systems using semi-markov kernel, IEEE Trans. Fuzzy Syst. 30 (7) (2022) 2289–2299.
[40] J. Wang, J. Xia, H. Shen, M. Xing, J.H. Park, H∞ synchronization for fuzzy Markov jump chaotic systems
with piecewise-constant transition probabilities subject to PDT switching rule, in: IEEE Transactions on Fuzzy
Systems, 29, 2021, pp. 3082–3092, doi:10.1109/TFUZZ.2020.3012761.
[41] X. Liu, J. Xia, J. Wang, et al., Interval type-2 fuzzy passive filtering for nonlinear singularly per-
turbed PDT-switched systems and its application, J. Syst. Sci. Complex 34 (2021) 2195–2218, doi:10.1007/
s11424- 020- 0106- 9.
[42] A. Farshidianfar, S. Soheili, Optimization of TMD parameters for earthquake vibrations of tall buildings includ-
ing soil structure interaction, Iran Univ. Sci. Technol. 3 (3) (2013) 409–429.
[43] C.C. Spyrakos, C.A. Maniatakis, I.A. Koutromanos, Soil–structure interaction effects on base-isolated buildings
founded on soil stratum, Eng. Struct. 31 (3) (2009) 729–737.
[44] S. Shourestani, F. Soltani, M. Ghasemi, S. Etedali, SSI effects on seismic behavior of smart base-isolated
structures, Geomech. Eng. 14 (2) (2018) 161–174.
[45] N. Aguila-Camacho, M.A. Duarte-Mermoud, Fractional adaptive control for an automatic voltage regulator, ISA
Trans. 52 (6) (2013) 807–815.
[46] I. Pan, S. Das, Frequency domain design of fractional order PID controller for AVR system using chaotic
multi-objective optimization, Int. J. Electr. Power Energy Syst. 51 (2013) 106–118.
[47] B.M. Vinagre, I. Podlubny, A. Hernandez, V. Feliu, Some approximations of fractional order operators used in
control theory and applications, Fract. Calculus Appl. Anal. 3 (3) (2000) 231–248.
[48] A. Oustaloup, La Commande CRONE: Commande Robuste D’ordre Non Entier, Hermes, 1991 ISBN:
2866012895 9782866012892.
[49] A. Oustaloup, X. Moreau, M. Nouillant, The CRONE suspension, Control. Eng. Pract. 4 (8) (1996) 1101–1108,
doi:10.1016/0967- 0661(96)00109- 8.

10562
S. Etedali, A.-A. Zamani, M. Akbari et al. Journal of the Franklin Institute 360 (2023) 10536–10563

[50] A. Oustaloup, J. Sabatier, X. Moreau, From fractal robustness to the CRONE approach, in: Proc., ESAIM, 5,
1998, pp. 177–192.
[51] A. Podlubny, Fractional-order systems and PIλ Dμ controllers, IEEE Trans. Autom. Control 44 (1) (1999) 208–
214, doi:10.1109/9.739144.

10563