09-10-2014, 04:24 PM
The protection of a building structure, the equipment housed within and the
human occupants against damage induced by large environmental loads e.g.,
earthquake, strong wind gusts and waves etc., is without doubt, a worldwide
priority. Requirements of safety of structures are addressed by the following
design considerations, namely, reliable information on service loads,
characteristics of the construction materials and efficiency and robustness of the
analysis and design procedure employed. The uncertainty involved in the
information about service loads, behavior and strength of the construction
materials and the approximations involved in the analysis and design methods
are such that these requirements can not be met completely.
Progress in material science and manufacturing technology has resulted in new
construction materials characterized by higher strength and predetermined
behavior, e.g., high strength structural steel and high performance concrete.
Better quality control norms in the manufacturing process have led to lesser
variability in the material proper-ties.
Availability of computationally powerful digital computers together with the use
of computationally powerful and accurate methods of structural analysis, e.g.,
finite element method, to a great extent and the evolution of mathematical tools
to handle uncertainties, e.g., probability theory, fuzzy sets and fuzzy logic, to
some extent, have led to the design and construction of complex, slender
structures such as high-rise buildings, long span bridges, deep water offshore
platforms, etc.
To ensure both the safety and serviceability of the structure during the entire
useful life of the structure, limit state design approaches have been developed.
In this approach partial factors have been associated with live, wind, earthquake
and dead loads, considering uncertainty levels in the peak magnitude of the
various loads. Likewise partial factors have been associated with the variability
in the strength of the materials or in other words resistance of the structure. This
approach has been further improved by incorporating other reliability measures
such as return period for environmental loads such as windstorm and
earthquakes.
As the structures have become more costly, more complex and serve more
critical functions, e.g., tall buildings, long span bridges, deep water offshore
platforms, nuclear power plants, etc., the consequences of their failure are
catastrophic. The safety of structures during the hazardous earthquake and
wind loads applies not only to the structure but also to the life safety of the
occupants. Hence, the conventional reliability criteria are no longer adequate.
To enhance the safety requirements, most of the buildings are currently designed
to yield during hazardous environmental loads without complete collapse, thus
protecting the occupants within the structure. Inelastic deformation (yielding) is
permitted as means of dissipating a portion of the energy transferred to the
structure by these environmental loads. However, plastic deformation due to
yielding results in permanent damage. Thus the conventional structural design
is often based on providing sufficient strength and stiffness to limit the inelastic
deformation to an acceptable level.
To meet the design requirements of economy, safety and serviceability, a new
design philosophy has emerged, which is based on specifying the expected
performance of the structural system during large environmental loads. Thus,
rather than designing to ensure only the life safety of the occupants, one may
ensure minimal damage to the structural system as well. The designs that follow
this alternate philosophy are known as performance based structural designs.
Structures cannot be designed to withstand all possible external loads, however,
some extraordinary loading episodes do occur, leading to damage or even failure
of the structure. Higher flexibility and low damping characteristics of slender
high rise structures have given rise to (i) higher material distress during large
environmental loads, such as earthquakes and wind gusts, resulting in the failure
of the structure and (ii) unacceptable levels of vibration causing discomfort to the
occupant.
One way to mitigate the effects of hazardous environmental loads and to meet
the stringent performance requirements is through the application of structural
control technology. The of structural vibrations can be accomplished by various
means, such as, providing counter forces (active control), isolating and
dissipating the energy of the excitations (base isolation), absorbing and
dissipating the energy of the structural vibrations (passive control). Structural
control can be implemented using either passive, active or combinations of these
strategies (semi-active, hybrid control).
Over the past three decades tremendous progress has been made to make control
a viable technology for enhancing structural functionality and safety against
natural hazards such as strong earthquakes and wind gusts. The passive and
active control systems have been used in real application to reduce the
earthquake and wind induced lateral vibrations of buildings.
Passive control systems operate without using any external energy supply.
These systems either use potential energy generated by the response of the
structure to supply the control force or dissipate the energy of the excitation
through friction or viscoelastic deformation. Passive control systems include
tuned mass dampers, tuned liquid dampers and a variety of energy dissipaters,
such as, base isolation systems against earthquake loads, metallic yield dampers,
friction dampers, viscoelastic Yi dampers, etc. Passive control system suffers
from their limited capability to control the structural response, as it does not use
any external energy source.
Active control systems operate using an external energy supply to apply the
control force on the structure. Active control systems use sensors to measure the
response of the system and/or excitation, compute the control command from
the sensor output using a control algorithm and apply the control command
to the structure by the means of actuators. Active mass dampers, active mass -
drivers, active tendon systems, etc., are some of the active control devices
developed and tested for actual structural application during the last two
decades. The active control system is dependent on an external power supply
and the power requirement is quite high in case of control of Civil Engineering
structures. This makes such system vulnerable to power failure, which is always
a possibility during a strong earthquake and windstorm. Moreover, due to the
high power requirement, it is difficult to provide an active control system with
its own dedicated power supply.
A hybrid control system is a viable solution to alleviate some of the limitations
that exist in either a passive or an active control system functioning alone. The
hybrid control system uses an active control component with a passive control
component to supplement and improve the performance of the passive control
system and to decrease the energy requirement of the active control system. In
case of power failure or failure of the active control component, the passive
component of the hybrid control system still offers some degree of protection,
making the system fail-safe, an essential design requirement for life-line
structures.
Active control system or the active control component of a hybrid control system
computes the control command from the sensor output using a control
algorithm. Performance of the control algorithms largely depends on the
accuracy of the model in handling the dynamics of the structure and control
devices. Uncertainties and nonlinearities in loads and material are inherent to
most structures, limiting the accuracy of the model of dynamics of the structure.
This rotation in the control algorithms can be alleviated using an intelligent
control system, e.g., logic control systems. An intelligent controller can be
designed without specifying a very precise and an accurate model of the
structural dynamics. Some of the characteristics of fuzzy logic appealing to
control engineers are its effectiveness and ease in handling nonlinearities,
uncertainties and heuristic knowledge.
Very few of the practical building structures have a zero eccentricity between the
axis of rotation and the axis of inertia and are subjected to pure unidirectional
excitation making them amenable for modeling as planar structures. A three
dimensional structure with an eccentric location of the axis of rotation and inertia
has coupled lateral and torsional responses, even when excited unidirectionally.
Symmetric buildings, i.e., the buildings that do not have any eccentricity between
mass and stiffness centres, respond with coupled lateral and torsional motion
under wind excitations having an eccentricity between the aerodynamic centre of
excitation and the mass \& stiffness centre of the building. However, twodimensional
plane frame structural models have been considered in most of the
recent research on structural control technology. Thus most of the structural
control methods that have been studied are not suitable for a majority of the
practical building structures.
Behavior of the structural material is not always linear and time invariant,
especially in the event of large environmental loads. A control system designed
for a linear, time invariant structural system may not perform effectively and
efficiently in the event of large environmental loads that lead to nonlinear
behavior of the structural material. Use of an adaptive control system that has a
capability of online tuning of the control system, based on the current dynamics
of the structure; can alleviate the limitation of the fixed type control system.
In the present study an approach for an optimal design of fuzzy logic controller
(FLC) driven active and hybrid control systems have been proposed. The
proposed approach has been applied to both the plane frame equivalent 2-D
models and torsionally coupled 3-D models of the building structure subjected to
earthquake and wind excitation. An effective and efficient arrangement of the
hybrid mass dampers has been presented for the torsionally coupled buildings.
The safety of the structure mainly depends on the displacement and rotational
response, while the comfort level of the occupants depends on the acceleration
response. To ensure both the structural safety and serviceability (occupants'
comfort), a multi-objective optimal design strategy has been formulated to
minimize displacement, acceleration and torsional (in case of torsionally coupled
building) responses of the structures. Genetic algorithm (GA) efficiently finds an
optimal solution from the complex and possibly non-convex discontinuous
solution space. A modified form of a two-branch tournament GA has been used
in the present study for multi-objective optimization of the control system
parameters, as the optimization problem is not necessarily convex.
The proposed approach has been further extended to design of an adaptive
control system to improve the robustness of the FLC for nonlinear building
structures. An online identification of the dynamics of the structure has been
incorporated using artificial neuro- fuzzy inference system (ANFIS). Optimal
changes in the FLC parameters have been computed for identified dynamics of
the structure. A "currently available" time history of the excitation and the set of
design input excitation have been incorporated in optimization problem to
establish the required changes in the parameters of the FLC. Various hardware
related constraints such as sensor noise, effects of quantization, saturation and
sampling time of the analog to digital converter (ADC) and digital to analog
converter (DAC), computational time delay to execute the main control
algorithm and the supervisory control algorithm in case of adaptive control
system, have been considered for a more realistic simulation of the digital
implementation of the control system.
Effectiveness of the proposed control strategies has been investigated for the
structural control benchmark problems and other problems for seismically and
wind excited building structures, available in the literature. To make the
comparison of the performance of the proposed control system with those that
are reported in the literature, respective model of the buildings and the
corresponding input excitations have been used in the present study. The
performance of the proposed control strategies has been found to be better than
the strategies reported in the literature. Multi-objective optimal design for FLC
based hybrid control system provides a set of 'Pareto- optimal’ designs that
enable the designer to select an appropriate control system design for the desired
performance requirements. Positions of the control devices and sensors have
been optimized with the other parameters of the control system. Position
optimization of control devices and sensors as a part of the controller design
optimization become necessary in the case of nonlinear structures where mode
shapes can not be used, to find out the optimal position of the control device.
The proposed control system is a hybrid control system having fail-safe
characteristic that makes the system suitable for lifeline structures where fail-safe
system is an essential design requirement.
The thesis has been organized in seven chapters and appendices encompassing
an introduction, a literature review, methodology and numerical results for fixed
and adaptive control system and conclusions. References cited in the thesis have