• 대한토목학회논문집
• /
• 제43권4호
• /
• pp.511-523
• /
• 2023

철근콘크리트 손상 감지를 위한 무인항공기와 딥러닝 연계에 대한 연구가 활발히 진행 중이다. 컨볼루션 신경망은 객체 분류, 검출, 분할 모델의 백본으로 모델 성능에 높은 영향을 준다. 사전학습 컨볼루션 신경망인 모바일넷은 적은 연산량으로 충분한 정확도가 확보 될 수 있어 무인항공기 기반 실시간 손상 감지 백본으로 효율적이다. 바닐라 컨볼루션 신경망과 모바일넷을 분석 한 결과 모바일넷이 바닐라 컨볼루션 신경망의 15.9~22.9% 수준의 낮은 연산량으로도 6.0~9.0% 높은 검증 정확도를 가지는 것으로 평가되었다. 모바일넷V2, 모바일넷V3Large, 모바일넷 V3Small은 거의 동일한 최대 검증 정확도를 가지는 것으로 나타났으며 모바일넷의 철근콘트리트 손상 이미지 특성 추출 최적 조건은 옵티마이저 RMSprop, 드롭아웃 미적용, 평균풀링인 것으로 분석되었다. 본 연구에서 도출된 모바일넷V2 기반 7가지 손상 감지 최대 검증 정확도 75.49%는 이미지 축적과 지속적 학습으로 향상 될 수 있다.

• 한국구조물진단유지관리공학회 논문집
• /
• 제15권3호
• /
• pp.171-177
• /
• 2011

본 논문에서는 비접촉식 충격-반향 기법을 이용한 콘크리트 구조물의 손상 탐지에 있어서 비용 효율성을 높이기 위한 다이나믹 마이크로폰의 적용 가능성에 대하여 알아보았다. 박리 손상이 인공적으로 모사된 콘크리트 슬래브 실험체에 대하여 비접촉 충격-반향 실험을 다이나믹 마이크로폰을 이용하여 수행하였으며 저비용 센서 시스템의 손상 탐지 성능을 분석하였다. 실험 결과 다이나믹 마이크로폰으로 의미있는 신호의 측정이 가능하며, 또한 콘크리트의 박리 손상도 고성능 음압 센서만큼 명확하게 검출 가능하다는 것을 알 수 있었다.

• International Journal of Advanced Culture Technology
• /
• 제10권3호
• /
• pp.393-396
• /
• 2022

Recently, rapid climate change has had a significant impact on the ecosystem of honeybees. In addition, the problem of Vespa Hornets invasion of colonies has a fatal impact on the bee ecosystem, independent of climate change. Especially in late summer. This study relates to a method for preventing Vespa Hornets attack. In this study, we developed a Vespa Hornets sound detection device was developed by collecting and analyzing the sound of a Vespa Hornets and applying IoT technology. The developed device detects the sound of a Vespa Hornets when Vespa Hornets appears around the hive of the bees and sends a signal to automatically close the door of the beehive. The device that receives the signal drives the motor that controls the honeycomb door to close the beehive door. The Vespa Hornets sound detection device operates until no Vespa Hornets sound is detected. The system developed by us is expected to be installed in the beehives of actual beekeeping farms to dramatically reduce the damage caused by by Vespa Hornets.

• Smart Structures and Systems
• /
• 제18권6호
• /
• pp.1189-1202
• /
• 2016

This paper describes an application of wavelet analysis for damage detection of a steel frame structure with bolted connections. The wavelet coefficients of the acceleration response for the healthy and loosened connection structure were calculated at each measurement point. The difference of the wavelet coefficients of the response of the healthy and loosened connection structure is selected as an indicator of the damage. At each node of structure the norm of the difference of the wavelet coefficients matrix is then calculated. The point for which the norm has the higher value is a candidate for location of the damage. The above procedure was experimentally verified on a laboratory-scale 2-meter-long steel frame. The structure consists of 11 steel beams forming a four-bay frame, which is subjected to impact loads using a modal hammer. The accelerations are measured at 20 different locations on the frame, including joints and beam elements. Two states of the structure are considered: healthy and damaged one. The damage is introduced by means of loosening two out of three bolts at one of the frame connections. Calculating the norm of the difference of the wavelet coefficients matrix at each node the higher value was found to be at the same location where the bolts were loosened. The presented experiment showed the effectiveness of the wavelet approach to damage detection of frame structures assembled using bolted connections.

• 한국항공우주학회지
• /
• 제37권11호
• /
• pp.1089-1095
• /
• 2009

태핑 테스트 시 방사되는 음압을 Rayleigh 적분식을 이용하여 해석하였다. 구조물에 의해 방사되는 음은 구조물의 거동과 직접적으로 연관되므로 충격응답해석이 수행되어야 한다. 본 논문에서는 층간분리가 존재하는 적층판의 해석을 위하여 층간분리 모델을 사용하였으며, 해머 형상의 충격체를 모사할 수 있는 효과적인 스프링-질량모델을 제시하였다. 예측된 음압이력과 시험결과와 비교하였으며, 음압이력과 충격하중이력에 손상이 주는 영향을 검토하였다. 방사되는 음압과 충격하중이력은 적층판에 존재하는 손상에 영향을 주는 것을 보였다. 결과적으로 제안된 음을 이용한 태핑실험은 복합재에 존재하는 손상을 검출하는데 신뢰성 있는 방법임을 알 수 있었다.

• 한국융합학회논문지
• /
• 제9권7호
• /
• pp.9-16
• /
• 2018

본 연구의 목적은 악의적인 사용자가 내부 출입을 위해 물리적인 충격으로 Door-Lock을 파손하는 경우가 있다. Door-Lock 시스템에 설정되어 있는 아날로그 변위 값과 일치하면 관리자 스마트폰에 실시간으로 파손 정보를 전송함으로써 신체 및 재산을 보호한다. 연구방법은 Door-Lock이 물리적인 충격으로 파손될 경우 등록된 사용자들에게 실시간으로 파손 정보를 전송하는 시스템을 제안한다. 그리고 Door-Lock에서 감지한 충격 정보와 감도조절부의 데이터를 비교한다. 제안한 시스템의 웹 서버에서는 Door-Lock에서 전송한 충격 정보를 DB에 저장한 후 충격 정보가 DB에 저장된 충격 감지 전송 기준 값보다 크면 SMS 모듈로 관리자에게 실시간으로 전송해서 악의적인 사용자의 불법적인 출입 정보를 확인한다.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2010년도 춘계학술대회 논문집
• /
• pp.636-637
• /
• 2010

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 1999년도 Proceedings of International Symposium on Remote Sensing
• /
• pp.243-248
• /
• 1999

From the lessons after the Nakhodka oil-spill in Jan. 1997, oil slick detection by using remote sensing data and assimilating the data to the simulation program is important for monitoring the oil-drift pattern. For this object, we are going to construct the oil-spill warning system for estimating the oil-drift pattern using remotesensing/numerical simulation Model. Additionally we plan to use this system for restorating oil-spill damage domestically, such as estimating the ecological damage and making the priority fur restorating the oil-spilled shoreline. This report is intended to summarize the role of geo-informatics in the oil spill accident by not only paying attention to the effect of information provision/information management via the map, but also reporting the interim result in part based on the details discussed in the processes of recovery support and environmental impact assessment during the Nakhodka's accident.

• 한국소음진동공학회논문집
• /
• 제15권9호
• /
• pp.1108-1117
• /
• 2005

A robust damage detection method using short-time Fourier transform and beating phenomena is presented as an estimating tool of the healthiness of large structures. The present technique makes use of beating phenomena that manifest themselves when two signals of similar frequencies are added or subtracted. Unlike most existing methods based on vibration signals, the present approach does not require an analytic model for target structures. Furthermore, the main advantage of the proposed method compared to the competing diagnostic method using vibration data is its robustness. The proposed method is not affected by the amplitude of exciting signals and the location of exciting points. From a measuring view point. the location of sensing point have no influence on the performance of the present method. With a view to verifying the effectiveness of this method. a series of experiments are made and the results show its possibility as a robust damage diagnostic method.

• Smart Structures and Systems
• /
• 제6권5_6호
• /
• pp.731-748
• /
• 2010

Impact detection and health monitoring are very important tasks for civil infrastructures, such as bridges. Piezoceramic based transducers are widely researched for these tasks due to the piezoceramic material's inherent advantages of dual sensing and actuation ability, which enables the active sensing method for structural health monitoring with a network of piezoceramic transducers. Wireless sensor networks, which are easy for deployment, have great potential in health monitoring systems for large civil infrastructures to identify early-age damages. However, most commercial wireless sensor networks are general purpose and may not be optimized for a network of piezoceramic based transducers. Wireless networks of piezoceramic transducers for active sensing have special requirements, such as relatively high sampling rate (at a few-thousand Hz), incorporation of an amplifier for the piezoceramic element for actuation, and low energy consumption for actuation. In this paper, a wireless network is specially designed for piezoceramic transducers to implement impact detection and active sensing for structural health monitoring. A power efficient embedded system is designed to form the wireless sensor network that is capable of high sampling rate. A 32 bit RISC wireless microcontroller is chosen as the main processor. Detailed design of the hardware system and software system of the wireless sensor network is presented in this paper. To verify the functionality of the wireless sensor network, it is deployed on a two-story concrete frame with embedded piezoceramic transducers, and the active sensing property of piezoceramic material is used to detect the damage in the structure. Experimental results show that the wireless sensor network can effectively implement active sensing and impact detection with high sampling rate while maintaining low power consumption by performing offline data processing and minimizing wireless communication.