This study aims to optimize the cochlea-inspired artificial filter bank (CAFB) using El-Centro seismic waveforms and test its performance through a shaking table test on a two-span bridge model. In the process of optimizing the CAFB, El-Centro seismic waveforms were used for the purpose of evaluating how they would affect the optimizing process. Next, the optimized CAFB was embedded in the developed wireless-based intelligent data acquisition (IDAQ) system to enable response measurement in real-time. For its performance evaluation to obtain a seismic response in real-time using the optimized CAFB, a two-span bridge (model structures) was installed in a large shaking table, and a seismic response experiment was carried out on it with El-Centro seismic waveforms. The CAFB optimized in this experiment was able to obtain the seismic response in real-time by compressing it using the embedded wireless-based IDAQ system while the obtained compressed signals were compared with the original signal (un-compressed signal). The results of the experiment showed that the compressed signals were superior to the raw signal in response performance, as well as in data compression effect. They also proved that the CAFB was able to compress response signals effectively in real-time even under seismic conditions. Therefore, this paper established that the CAFB optimized by being embedded in the wireless-based IDAQ system was an economical and efficient data compression sensing technology for measuring and monitoring the seismic response in real-time from structures based on the wireless sensor networks (WSNs).

본 연구에서는 압축센싱 기술인 CAFB(달팽이관 원리기반의 인공필터뱅크)가 El-Centro 지진 등 지진상황에서 구조물의 동적응답을 포함한 지진응답을 실시간으로 압축하여 획득할 수 있는지 평가하였다. 최적화된 CAFB를 무선 IDAQ 시스템에 임베디드 하였다. 이를 이용하여 대형 Shaking Table에 설치된 2-span 교량구조물의 지진응답을 실시간으로 압축하여 획득 하였다. 연구결과 압축신호는 원시신호 대비 우수한 응답성능 및 데이터 압축효과를 보였으며 이를 통해 CAFB가 지진상황에서도 구조물의 동적응답을 포함한 지진응답을 실시간으로 압축 획득할 수 있음을 확인하였다.

데이터 병목현상 등의 이유로 대용량의 동적 데이터를 무선으로 계측하여 구조물 건전도 모니터링을 수행하는 데는 한계가 있다. 이에 선행연구에서는 압축센싱 기술인 CAFB(달팽이관-영감형 인공필터뱅크)를 개발하여 무선 압축센싱 성능을 실험적으로 평가하였다. 본 논문에서는 CAFB를 적용한 무선 계측시스템의 유효성 평가를 목적으로 Kobe 지진파형으로 최적화된 CAFB를 이용해 지진응답 실험을 진행하였다. 결과적으로 Kobe 지진파형으로 최적화한 CAFB로 수신한 무선 데이터는 64~91%의 압축률을 보였으며 99% 이상의 재건율로 무선 데이터를 획득하는데 효과적임을 확인하였다.

This paper aims at featuring dynamic characteristic of 2-span bridge by means of an cochlea-inspired artificial filter bank(CAFB), for the purpose of effective acquirement of real-time valid dynamic(acceleration) responses. Through dynamic testings performed on a model of 2-span bridge, its dynamic characteristics were featured on natural frequency. They were evaluated in comparison to wired measurement responses. Finally, the dynamic wireless measurement system on the basis of CAFB were found technologically and economically effective by obtaining valid dynamic(acceleration) responses whose size was compressed over the frequency range of interest.