Optimal placement of sensors on buildings subjected to intermediate‐storey excitation according to QN control method

Abstract

SUMMARYAn algorithm to calculate direct velocity feedback gain with limited number of sensors is developed in a simple way such that a certain performance index is minimized according to QN control method. If a limited number of velocity outputs can be measured, full velocity responses of the whole structure can be interpolated based on the mode shapes. By deÿning the performance index function as a combination of the structure's velocity responses and control forces only, feedback gain can be determined according to QN control method with the external excitation being taken into account throughout the entire algorithm. Control forces are then regulated by the time-invariant feedback gain matrix. The e ective location of the active control devices for a building structure subjected to intermediate-storey excitation has been determined to be in the three oors adjacent to the vibration source. Hence for the purpose of this paper, only the optimal placement of sensors is veriÿed. It is shown in this paper that if the dynamic behaviour of the structure is well described by a mathematical model, su cient response reduction e ect can be achieved according to the new DVFC algorithm, and the degradation of control performance due to time delay can also be veriÿed.

Structural seismic response reconstruction method based on multidomain feature-guided generative adversarial neural networks

“…In this section, the sensor optimization arrangement scheme for an 8-story frame structure is discussed, with the accuracy of the MWGAN-TF reconstructed response as an orientation. The goal of sensor optimization is to obtain better response reconstruction with N sensors arranged on M = 8 nodes, and all the arrangement schemes are C 1 8 × 2 + C 2 8 × 2 + C 3 8 × 2 + C 4 8 × 1 = 255 in total. In each scheme, the response data of the nodes with arranged sensors were used as input data, the response data of the nodes without arranged sensors were used as output data, respectively, and the corresponding structural response reconstruction models were built.…”

“…However, the number of sensors installed on structures is always limited due to economic costs, test instrumentation and site conditions. In practice, sensors are typically installed only in a few optimized critical locations [3,4]. As a result, most structures equipped with SHM systems are not sufficiently sensitized.…”