Comparative Study on Multi-Objective Genetic Algorithms for Seismic Response Controls of Structures - Design Optimization of Active and Passive Structural Control Systems - page 351

Figure 12. Maximum drift ratio responses (El-Centro, Hachinohe, Northridge, and Kobe earthquakes)

of high-rise buildings. In near future, sensitivity
analyses of the sensor fault (Sharifi et al. 2010)
and structural damage to the performance of the
optimal control devices layouts in high-rise build-
ings will be investigated.

the IRR-GA and a strength Pareto evolutionary
algorithm 2 (SPEA 2). The last one is Gene Ma-
nipulation GA (GMGA) that is developed based
on novel recombination and mutation mechanism.
The MOGAs are formulated as optimization prob-
lems of finding optimal locations and number of
actuators and sensors within seismically excited
large-scale civil structures such that dynamic
responses of structures are also minimized. To
implement active control systems into seismi-
cally excited structures, linear quadratic regulator
(LQR)-based controllers, Kalman estimators,
hydraulic actuators, and accelerometers are used.
To demonstrate the effectiveness of the proposed
three MOGAs, twenty-story two dimensional (2D)
and three dimensional (3D) building models are
developed using the finite element method. To ex-
cite those large-scale building models, a variety of
earthquakes are used as external loads. Further, the
performances of the three MOGAs are compared
in terms of convergence rate, Pareto fronts, the
time history responses, and maximum interstory
responses. It is shown from the simulations that
the proposed MOGAs are very effective in finding
not only optimal locations and numbers of actua-

CONCLUSION

To date, most encoding policies of genetic algo-
rithms (GAs) are based on binary or real-coded
encoding. However, these encoding policies may
not be appropriate for solving highly complex and
computationally intensive optimization problems,
e.g., large-scale infrastructure design and analysis
problems. To address the issue, this topic chapter
presents three novel frameworks of multi-objec-
tive genetic algorithms (MOGAs) for integrated
optimal design of actively controlled large-scale
infrastructures under seismic excitation, by com-
bining the best features of several GAs. The first
MOGA is developed through the integration of an
implicit redundant representation (IRR) genetic
algorithm (GA) and a non-dominated sorting II
(NS2) GA. The second is proposed by combining