Development and validation of a novel earthquake damage estimation scheme based on the continuous wavelet transform of input and output acceleration measurements

This paper presents the development and validation of several novel data-driven damage sensitive features. The proposed features are based on the Continuous Wavelet Transform of both the input acceleration signal to the structure and the output acceleration response. The combination of the input and output wavelet coefficients and the derivation of the features is presented. The features are applied to experimental data obtained from shake table tests on reinforced concrete bridge columns. The results are compared against typically used damage metrics, such as hysteretic energy, and exhibit high correlation with damage. The performance of the features in binary damage detection is evaluated using numerical simulations of reinforced concrete columns under earthquake loading. A damage classification scheme based on the developed features and established damage indices is proposed and validated through Monte Carlo simulation. The proposed features are applied to experimental and simulated data from a multistory frame, illustrating the features' capabilities for damage localization in civil structures. Due to its data-driven nature and use of strong motion recordings, the proposed damage detection scheme can be tailored to a wide variety of applications and deliver damage information immediately after an earthquake. Copyright (c) 2014 John Wiley & Sons, Ltd.