Premium
Experimental study of identification and control of structures using neural network. Part 1: identification
Author(s) -
Bani–Hani Khaldoon,
Ghaboussi Jamshid,
Schneider Stephen P.
Publication year - 1999
Publication title -
earthquake engineering and structural dynamics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.218
H-Index - 127
eISSN - 1096-9845
pISSN - 0098-8847
DOI - 10.1002/(sici)1096-9845(199909)28:9<995::aid-eqe851>3.0.co;2-8
Subject(s) - system identification , artificial neural network , engineering , actuator , time domain , frame (networking) , identification (biology) , control system , earthquake engineering , vibration , frequency domain , structural engineering , control theory (sociology) , control engineering , simulation , computer science , control (management) , artificial intelligence , acoustics , mechanical engineering , data modeling , botany , physics , electrical engineering , software engineering , computer vision , biology
Experimental verifications of a recently developed structural control method using neural network has been carried out on the earthquake simulator at the University of Illinois at Urbana—Champaign. The test specimen was a 1/4 scale model of a three‐storey steel frame. The control system consisted of a tendon/pulley system controlled by a single hydraulic actuator. The model structure had a total mass of 2994 kg (6600 lb), distributed evenly among the three floors, and a total frame height of 254 cm (100 inches). The structure had three distinct lightly damped fundamental modes of vibration plus two higher modes representing the structure–control interaction and the actuator dynamics. The system identification and parameter estimation have been conducted in two experimental methods: first, the system has been identified in the time domain and the estimated parameters were used in the frequency domain methods and secondly, the system was modelled and identified using multiple emulator neural networks with different prediction capabilities. These emulators were employed in the control design. This paper describes the test set‐up, the experimental validation of the identified model in the time and frequency domains, and experimentally demonstrates the performance of the multiple emulator neural networks. Copyright © 1999 John Wiley & Sons, Ltd.