AN UPDATED TIME DOMAIN SOFT PARAMETER ESTIMATION METHODOLOGY FOR A FRAME STRUCTURE UNDER BASE EXCITATION
In recent years, neural networks-based identification method using vibration measurement time series without any mode shapes and frequency extraction have been proposed. In this paper, an updated time domain soft structural parameter estimation methodology for multi degree-of-freedom structure using displacement response and excitation acceleration measurements is proposed. A displacement based emulator neural network (DENN) for displacement forecasting of a reference structure and a parametric evaluation neural network (PENN) for structural inter-story stiffness estimation are constructed to facilitate the whole identification process. Based on the trained DENN and the PENN, the inter-storey stiffness of the object structure is identified using an evaluation index called root mean square of prediction difference vector (RMSPDV). The performance of the proposed methodology is examined with simulated relative displacement measurements of a 2-story frame model structure under base excitation. Results show that the proposed neural- network-based methodology may be an applicable method for structural damage detection.
Bin Xu Ping Lu Gang-Bing Song
College of Civil Engineering, Hunan University, Chang sha 410075, P.R. China Department of Mechanical Engineering, University of Houston, USA
国际会议
长沙
英文
1490-1496
2007-11-19（万方平台首次上网日期，不代表论文的发表时间）