Identification of nonlinear hysteretic systems using sequence model‐based optimization. Issue 4 (22nd January 2020)

• Record Type: Journal Article
• Authors: Jiang, Kejie; Wen, Jianian; Han, Qiang; Du, Xiuli
• Is Part Of: Structural control and health monitoring. Volume 27:Issue 4(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-01-22
• Subjects: Bouc–Wen model -- hyperparameter optimization -- nonlinear hysteretic system -- parameter identification -- sequence model optimization; Structural engineering -- Periodicals; Structural control (Engineering) -- Periodicals; Automatic data collection systems -- Periodicals; Detectors -- Periodicals; 624.17
• DOI: 10.1002/stc.2500
• Languages: English
• ISSNs: 1545-2255

Abstract

Summary: Identification of nonlinear hysteretic systems has practical significance for the prediction of structural response. This paper develops a novel parameter identification method for nonlinear hysteretic systems. The Bouc–Wen model and its improved models are used to parametrically characterize the structural systems. Model parameters are identified using a given load‐displacement trajectory. The proposed method is developed under the framework of the sequence model optimization, which provides a broader search domain and needs fewer iterations. Firstly, a series of sensitivity analyses were conducted to investigate the effects of the Bouc–Wen parameters on the hysteresis behavior and to guide the identification process. Subsequently, based on the framework of the sequence model optimization, a novel strategy for identifying the model parameters of the nonlinear hysteresis system was proposed. Finally, the effectiveness and accuracy of the proposed algorithm were verified by the experimental data and numerical simulation results. The results indicate that the identified numerical model can accurately capture the strength and stiffness degradation and the pinching effect of the structural system. As a result, the predicted hysteretic trajectories are well agreed with the actual structural responses. The case studies demonstrate that the proposed method is sufficiently accurate and computationally efficient for the nonlinear hysteretic systems.