Structural health monitoring techniques implemented on IASC–ASCE benchmark problem: a review

Various structural health monitoring techniques have been developed over the years. Due to the lack of a common platform to test the efficiency of these methods, the damage analysis models have been tested on different structures selected according to the choice of researches. Therefore, perfect comparison among the models was not possible. In light of this event, a benchmark structure was developed providing a common ground to analyse the effectiveness of the damage detection strategies. This structural damage analysis consists of different damage patterns, major damages and minor damages. The damage detection algorithms were tested for the detection of these different damage patterns and the effectiveness against noise contamination. Also the amount of data required for the algorithms to effectively detect damage was also recorded, which indicated the efficiency of the method applied. The paper deals with the application of different damage detection techniques on the ASCE benchmark Phase-I and Phase-II structure and studies their efficiency against the other structures. A brief comparison has been made among different damage detection models such as Bayesian model, neural network, autoregressive models, and model update. These methods have been successfully implemented on the benchmark structure and their efficiencies have been measured in terms of noise contamination, the amount of data required to measure the damage and the detection of damage (major and minor). Out of all the techniques discussed, model update technique, wavelet approach, autoregressive technique, convolution neural network and synchrosqueezed wavelet transform have proved to a robust damage analysing tool.