Neuromorphic Smart Controller for Seismically Excited Structures - Design Optimization of Active and Passive Structural Control Systems - page 218

the emotional learning and process. However, as
a macro-level mechanism, the emotional learning
process is at a much higher level than the ANNs
that consist of a very large number of neurons. In
other words, it is not necessary to apply an ANN
with high-cost computational loads to system level
designs. It can be inferred from this perspective
that an emotion-based learning algorithm can be
easily implemented without high-computational
loads, compared to the ANNs. However, relatively
little research has been carried out on an emotion-
based control mechanism due to the fact that: (1)
the properties and mechanisms of emotions in the
human brain are not clearly understood and (2) a
mathematical model of the emotional learning and
process mechanisms in the human brain is only
beginning to be developed (Balkenius & Moren
2001; Moren 2002).
Moren and Balkenius (2000) proposed a
mathematical model of emotional learning pro-
cess that occurs in human brain describing the
physical phenomenon of the emotional processing.
Since then, investigators have applied the brain
emotional learning (BEL) algorithm to feedback
control problems: Lucas et al. (2004) introduced
the BEL algorithm for control system design;
Chandra and Langari (2006) investigated the
stability issues of the BEL algorithm; Mehrabian
et al. (2006) used the BEL algorithm for a flight
control system design by eliminating tracking
errors without prior knowledge of the plant
dynamics; Mehrabian and Lucas (2006) also ap-
plied the BEL algorithm to various benchmark
nonlinear dynamic systems; Shahmirzadi et al.
(2006) proved that the brain limbic system can be
applied to a 14-DOF model of a tractor-semitrailer;
Sheikholeslami et al. (2006) achieved the adaptive
set point control and disturbance rejection of an
HVAC system using the BEL control algorithm;
based on the BEL algorithm, Rouhani et al. (2006)
solved the output temperature tracking problem

of the electrically heated micro-heat exchanger;
Rouhani et al. (2007) also showed the excellent
performance of the BEL controller for the rotor
speed and position of a switched reluctance mo-
tor; the control performance of the BEL controller
is also experimentally verified by Jamali et al.
(2009) using a digital pendulum system; Kim and
Langari (2009, 2011) proposed the BEL control-
ler based mobile robot target tracking method;
more recently, Kim and Langari (2010a, 2010b)
developed autonomous vehicle functions such as
lane change maneuver and adaptive cruise control
by BEL control strategy. They also compared the
control results with conventional control methods,
i.e. Fuzzy, PID, and human driver model and
showed the robustness and performance of the
proposed controller. In the previous researches
of the authors (Kim and Langari 2009, Kim et
al. 2010, Kim and Langari 2010, and Kim and
Langari 2011), the authors compared the control
performances of the neuromorphic controller with
those of conventional control methods such as
passive, PID, PD, LQG, and fuzzy logic control
method. From the comparisons it is observed
that the main advantages of the neuromorphic
smart control are the robustness to the parameter
uncertainties, error elimination, and fast response.
In this article, a new control algorithm for
seismic response control of building structure-
magnetorheological (MR) systems is proposed.
The control algorithm is developed through
the integration of the BELBIC algorithm with
a proportional-integral-derivative (PID) and a
semiactive inversion algorithm. This chapter
is organized as follows: Section 2 presents the
proposed neuromophic control algorithm. Smart
structures, i.e., building-magnetorheological
damper systems, are described, including simula-
tion results, in Section 3. Concluding remarks are
discussed in Section 4.