Optimal placement of active control devices and sensors in frame structures using multi‐objective genetic algorithms. Issue 1 (11th August 2011)

• Record Type: Journal Article
• Authors: Cha, Young‐Jin; Raich, Anne; Barroso, Luciana; Agrawal, Anil
• Is Part Of: Structural control and health monitoring. Volume 20:Issue 1(2013:Feb.)
• Page Start: 16
• Page End: 44
• Publication Date: 2011-08-11
• Subjects: active control -- sensors -- structural control -- optimization -- algorithms -- steel frames; Structural engineering -- Periodicals; Structural control (Engineering) -- Periodicals; Automatic data collection systems -- Periodicals; Detectors -- Periodicals; 624.17
• DOI: 10.1002/stc.468
• Languages: English
• ISSNs: 1545-2255

Abstract

SUMMARY: Active control efficiency is highly dependent on the control algorithm and device types as well as the locations of the devices and sensors in a building. A gene manipulation, multi‐objective genetic algorithm is proposed to optimize the placement of active devices and sensors in frame structures to reduce active control cost and increase the structural control strategy's effectiveness. Gene manipulation uses engineering judgment to modify the encoded variable information defining the number of devices and sensors per floor in selected Pareto‐optimal front individuals. The proposed methodology evolves Pareto‐optimal layouts that minimize the number of devices/sensors used while also minimizing the building interstory drift for a 20‐story steel‐frame building under earthquake loading. The results indicate that the number and location of the devices and sensors in the layouts obtained strongly depends on the desired maximum drift. Also, the location of the sensors significantly impacts the efficiency of the active controller in reducing interstory drifts. In simulation trials, the proposed gene manipulation method obtained layouts that distributed devices and sensors more evenly over the building height than layouts obtained using standard multi‐objective methods, resulting in greater control efficiency. The primary benefit of implementing the proposed gene manipulation was in reducing the number of multi‐objective genetic algorithm generations required by up to 40% without negatively impacting the quality of Pareto‐optimal device/sensor layout solutions obtained. Copyright © 2011 John Wiley & Sons, Ltd.