Multi‐level design model and genetic algorithm for structural control system optimization

Li, Q. S.; Liu, D. K.; Zhang, Nong; Tam, C. M.; Yang, Lin

doi:10.1002/eqe.48

Cited by 23 publications

(8 citation statements)

References 15 publications

(18 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Search algorithms may directly be utilized by pairing the technique with an appropriate cost function (see, for example, [29][30][31]). Genetic algorithms are also commonly used for this application [6,7,[32][33][34]. Although this approach can be integrated with the control design process to improve performance of the resulting design [1], it can be time consuming.…”

Section: Placement Of Control Devicesmentioning

confidence: 99%

Smart system design for a 3D base-isolated benchmark building

Wang

Dyke

2008

Struct. Control Health Monit.

View full text Add to dashboard Cite

The application of control technologies to structures is expected to enhance a structure's performance in response to natural hazards. This study focuses on the development of control design strategies suitable for smart base isolation systems. The phase I benchmark problem for 3D base-isolated buildings is used as a case study to facilitate a direct comparison with the results of other designs. The smart base isolation system consisting of passive isolators and controllable semiactive devices at the base is employed for the demonstration and validation of these strategies. Magnetorheological (MR) dampers are used as the control devices and are placed in parallel with the isolation system. An H 2 /LQG control algorithm is used to determine the nominal control action and is used in conjunction with a clipped optimal decision block. Effective design for this complex system requires the use of control device and sensor placement techniques achieved with the consideration of the structure's physical properties. The use of correlation is also proposed to minimize redundancies in the placement of the sensors and devices. Simulation results demonstrate that the control technique is effective in improving the structural performance for a variety of earthquakes. Normalized inter-story driftJ 5 ðqÞ ¼ max t;f jja f ðt; qÞjj max t;f jj# a f ðt; qÞjj 3. Normalized floor acceleration J 4 ðqÞ ¼ max t;f jjd f ðt; qÞjj max t;f jj # d f ðt; qÞjj 4. Normalized RMS base drift J 7 ðqÞ ¼ max i jjs d ðt; qÞjj max i jjs # d ðt; qÞjj 5. Normalized RMS inter-story drift J 11 ðqÞ ¼ max f jjs d f ðt; qÞjj max f jjs # d f ðt; qÞjj 6. Normalized RMS floor accelerations J 8 ðqÞ ¼ max f jjs a ðt; qÞjj max f jjs # a ðt; qÞjj Y. WANG AND S. DYKE 942

show abstract

Section: Placement Of Control Devicesmentioning

confidence: 99%

Smart system design for a 3D base-isolated benchmark building

Wang

Dyke

2008

Struct. Control Health Monit.

View full text Add to dashboard Cite

show abstract

“…Li et al [14] propose a multilevel genetic algorithm in which active tendons are designed to control the top storey displacements in tall buildings due to wind and earthquake excitations. The optimization is considered in three levels that determine the number of actuators, the location of actuators and the optimal control gains.…”

Section: Introductionmentioning

confidence: 99%

Evolutionary aseismic design and retrofit of structures with passive energy dissipation

Dargush

Sant

2005

Earthquake Engng Struct. Dyn.

View full text Add to dashboard Cite

SUMMARYA new computational framework is developed for the design and retroÿt of building structures by considering aseismic design as a complex adaptive process. For the initial phase of the development within this framework, genetic algorithms are employed for the discrete optimization of passively damped structural systems. The passive elements may include metallic plate dampers, viscous uid dampers and viscoelastic solid dampers. The primary objective is to determine robust designs, including both the non-linearity of the structural system and the uncertainty of the seismic environment. Within the present paper, this computational design approach is applied to a series of model problems, involving sizing and placement of passive dampers for energy dissipation. In order to facilitate our investigations and provide a baseline for further study, we introduce several simpliÿcations for these initial examples. In particular, we employ deterministic lumped parameter structural models, memoryless ÿtness function deÿnitions and hypothetical seismic environments. Despite these restrictions, some interesting results are obtained from the simulations and we are able to gain an understanding of the potential for the proposed evolutionary aseismic design methodology.

show abstract

“…Abdullah (2001) [24] combined GA and a gradient-based optimization technique for optimal placement of controllers in a direct velocity feedback control system. Li (2001) [25] have developed a multilevel GA to solve a multitasking optimization problem. Singh et al (2002) [26] considered the placement of passive devices in a multistory building.…”

Section: Introductionmentioning

confidence: 99%

Evolutionary Constrained Design of Seismically Excited Buildings: Sensor Placement

Rowhanimanesh

Khajekaramodin

Akbarzadeh-T

2009

Advances in Intelligent and Soft Computing

View full text Add to dashboard Cite

Abstract. Appropriate sensor placement can strongly influence the control performance of structures. However, there is not yet a systematic method for sensor placement. In this paper, a general method based on a proposed constrained GA is suggested to optimally place sensors in structures. The optimal placement scheme is general for passive, active and semi-active controls and it does not depend on the control strategy and nonlinear dynamics of the control system. Due to low cost, high reliability, control effectiveness as well as installation simplicity, acceleration type sensors are considered. The proposed method is applicable to new or existing buildings, as the accelerometer placement is trivial. The efficiency of proposed method is evaluated on the benchmark building for placement of 5 and 3 sensors. The results show that the performance of control system with 5 and even 3 optimally placed sensors is at least 8% better than the original benchmark building design with 5 sensors. Generally, the proposed method is a simple and efficient practical approach to achieve improved performance using the fewest number of instruments.

show abstract

Multi‐level design model and genetic algorithm for structural control system optimization

Cited by 23 publications

References 15 publications

Smart system design for a 3D base-isolated benchmark building

Smart system design for a 3D base-isolated benchmark building

Evolutionary aseismic design and retrofit of structures with passive energy dissipation

Evolutionary Constrained Design of Seismically Excited Buildings: Sensor Placement

Contact Info

Product

Resources

About