본 논문에서는 유지보수를 위한 구조물 인상 시 위험도 분석을 수행하여 안전사고를 방지할 수 있는 스마트 자동 인상 시스템을 개 발하였다. 쌍대비교행렬 분석 기법을 활용하여 위험도를 분석할 수 있는 정량적 위험도 분석 프로그램을 개발하였고, 이를 자동 인상 시스템과 연계하여 구조물 인상과 동시에 실시간으로 위험도 분석을 하였다. 자동 인상 시스템의 구성요소 중 거리측정센서로 구조 물 인상 시의 변위를 측정하고, 측정된 변위는 정량적 위험도 분석 프로그램에 입력되어 위험도를 분석한다. 개발한 스마트 자동 인상 시스템의 성능을 확인하기 위해 실제 교량을 대상으로 실험을 수행하였으며, 구조물 인상과 동시에 위험도 분석이 가능한지를 확인 하였다. 스마트 자동 인상 시스템의 성능을 평가하기 위해 인상실험 시 검증된 LVDT(linear variable differential transformer)를 함께 설치하였으며 거리측정센서와 LVDT로 측정되는 변위로 최대 인상량과 구역별 단차를 분석하였다. 인상장치의 동시 작동에 대한 성 능을 통계적 분석방법인 분산분석(analysis of variance) 방법을 이용하여 성능을 검증하였다.

Pest control treatment was carried out using an unmanned automatic pesticide spraying system that can spray pesticides on crops while moving autonomously to control pests in vegetable greenhouse. As a result of examining the control effect on tomato and strawberry on thrips (Frankliniella occidentalis) and greenhouse whitefly (Trialeurodes vaporariorum) pests, 85.6% of yellow flower thrips were found in tomatoes and 87.5% in strawberries, and 81.7% (tomato) and 80.6% (strawberry) of greenhouse whitefly. In addition, the control effect according to the pesticide treatment method showed a control effect of 81.7% of the chemical spraying treatment by manpower and 83.9% of the automatic moving pesticide spraying treatment (F=22.1, p < 0.001). When comparing the control effect between the two treatment sections, there was no significance, but the automatic transfer spraying treatment showed a 2.2% higher effect. On the other hand, as a result of comparing the spraying time of the drug, the automatic unmanned control sprayer had a spraying time of 5 min/10a, which took about 25 min less than the conventional manpower spraying time of 25－30 min/10a. Based on these results, it was judged that the automatic transfer spraying method could be usefully used for efficient pest control in the facility greenhouse during the peak period of development.

Jagyeokru, an automatic striking water clock described in the Sejong Sillok (Veritable Records of King Sejong) is essentially composed of a water quantity control device and a time-signal device, with the former controlling the amount or the flow rate of water and the latter automatically informing the time based on the former. What connects these two parts is a signal generating device or a power transmission device called the ‘Jujeon’ system, which includes a copper rod on the float and ball-racked scheduled plates. The copper products excavated under Gongpyeong-dong in Seoul include a lot of broken plate pieces and cylinder-like devices. If some plate pieces are put together, a large square plate with circular holes located in a zigzag can be completed, and at the upper right of it is carved ‘the first scheduled plate (一箭).’ Cylinder-like devices generally 3.8 cm in diameter are able to release a ball, and have a ginkgo leaf-like screen fixed on the inner axis and a bird-shaped hook of which the leg fixes another axis and the beak attaches to the leaf side. The lateral view of this cylinder-like device appears like a trapezoid and mounts an iron ball. The function of releasing a ball agrees with the description of Borugak Pavilion, where Jagyeokru was installed, written by Kim Don (1385 ~ 1440). The other accounts of Borugak Pavilion’s and Heumgyeonggak Pavilion’s water clocks describe these copper plates and ball releasing devices as the ‘Jujeon’ system. According to the description of Borugak Pavilion, a square wooden column has copper plates on the left and right sides the same height as the column, and the left copper plate has 12 drilled holes to keep the time of a 12 double-hours. Meanwhile, the right plate has 25 holes which represent seasonal night 5-hours (Kyeong) and their 5-subhours (Jeom), not 12 hours. There are 11 scheduled plates for seasonal night 5-hours made with copper, which are made to be attached or detached as the season. In accordance with Nujutongui (manual for the operation of the yardstick for the clepsydra), the first scheduled plate for the night is used from the winter solstice (冬至) to 2 days after Daehan (大寒), and from 4 days before Soseol (小雪) to a day before the winter solstice. Besides the first scheduled plate, we confirm discovering a third scheduled plate and a sixth scheduled plate among the excavated copper materials based on the spacing between holes. On the other hand, the width of the scheduled plate is different for these artifacts, measured as 144 mm compared to the description of the Borugak Pavilion, which is recorded as 51 mm. From this perspective, they may be the scheduled plates for the Heumgyeonggak Ongru made in 1438 (or 1554) or for the new Fortress Pavilion installed in Changdeokgung palace completed in 1536 (the 31st year of the reign of King Jungjong) in the early Joseon dynasty. This study presents the concept of the scheduled plates described in the literature, including their new operating mechanism. In addition, a detailed model of 11 scheduled plates is designed from the records and on the excavated relics. It is expected that this study will aid in efforts to restore and reconstruct the automatic water clocks of the early Joseon dynasty.

PURPOSES : In this study, a Design Support Automatic Program for NATM (new Austrian tunneling method) tunnels based on BIM that can generate shape information including property information via the development of BIM libraries and the construction of a library design support automatic program is developed. METHODS : In this study, the abovementioned program is developed as an add-in Revit program. AutoDesk's program has the highest utilization rate, and it allows support tools to be added easily. Construction information classification based on WBS(Work Breakdown Structure) is developed. Subsequently, library construction and program development algorithms are developed. RESULTS : The automatic program constructs a library that reflects geometric characteristics through the construction of a library automatic module as well as generates three0-dimensional (3D) shape information by mapping between the constructed library and 3D informal alignment. The proposed program can automatically assemble a constructed library through the development of an algorithm for generating sequences of 3D shape information and automatically place models on a 3D terrain. CONCLUSIONS : The proposed automatic program can be designed and changed easily. It is expected to contribute to improvements in the speed of 3D NATM tunnel modeling and the efficiency of NATM tunnel design.

In this research, we have developed the non-contact automatic washer system for two-wheel motor vehicle through structural and fluid analysis. The structural analysis of mainframe was performed for the safety of the bike washer system, and multi-phase fluid analysis was applied for the washer cleaning performance which was analyzed the surface shear stress while varying the nozzle angle. As a result, we have developed the washer system by deriving the spray angle for the shear stress area and by analyzing the housing shape that takes into consideration structural stability. The performance of this washer system was verified by measuring the gloss before and after cleaning in the evaluation of the cleaning performance test of the system.

본 연구는 다양한 차종의 승용 자동 변속기 변속 성능 및 내구성 평가를 위한 시험 장비 설계를 위하여 수행되었다. 다양한 차종의 주행 중 변속 특성을 시험 장비에서 구현하기 위하여 가장 작은 배기량을 갖는 차량과 가장 큰 배기량을 갖는 차량의 엔진 성능 곡선, 차량 총 중량, 총 기어비를 분석하였다. 분석된 결과를 바탕으로 하여 차량의 총 중량에 의한 관성이 변속기 후단에 전달되었을 때의 등가 관성을 각각 계산하였다. 다양한 차종의 엔진 관성과 성능 곡선을 만족시키기 위해서 입력 동력원은 정격 출력 440㎾, 정격 토크 146.7 kg·m, 관성모멘트 1.76 kg·m2으로 선정한 후 차종에 따라 3가지의 증속기(2.1:1~4.2:1)를 선택할 수 있으며 추가적인 기계 관성을 증속기 후단에 장착할 수 있도록 설계한다. 출력 동력원은 정격 출력 520㎾, 정격 토크 500kg·m, 관성모멘트 5kg·m2로 선정한 후 2:1 감속기와 추가적인 기계 관성을 장착할 수 있도록 설계한다. 다양한 차종의 자동 변속기 변속 특성 및 내구성을 간단한 장치 변경을 통하여 하나의 시험 장비에서 시험할 수 있는 방법을 대상 차량의 엔진 및 관성 정보를 분석하여 제안하였다. 이러한 시험 장비를 활용한다면 자동변속기 개발 및 성능 향상을 위한 비용 및 시간을 크게 절감할 수 있을 것이라 판단된다.

본 논문에서는 생성 모델로 색칠하기 게임에서 사용 가능하도록 임의의 선화와 원하는 컬러링 스타일을 입력하면 자동으로 컬러링 영상을 생성하는 신경망 모델인 FillingGAN을 제안한다. 제안된 모델은 스타일 영상의 특징을 추출하는 오토 인코더 구조의 모듈과 추출된 스타일 영상의 특징을 선화에 적용해서 이미지를 생성하는 GAN 모델로 구성된다. GAN 모델은 선화에서 추출된 구조와 스타일 영상에서 추출된 색 정보를 이용해서 채색 영상을 생성하는 과정을 수행하며, 이를 위해서 선화의 구조와 스타일 영상의 색 정보를 유지하는 손실 함수를 설계한다. 우리의 모델은 선화의 고유한 특징을 보존하며 스타일이 적용된 이미지를 생성한다.

PURPOSES : This study uses deep learning image classification models and vehicle-mounted cameras to detect types of pavement distress — such as potholes, spalling, punch-outs, and patching damage — which require urgent maintenance. METHODS : For the automatic detection of pavement distress, the optimal mount location on a vehicle for a regular action camera was first determined. Using the orthogonal projection of obliquely captured surface images, morphological operations, and multi-blob image processing, candidate distressed pavement images were extracted from road surface images of a 16,036 km in-lane distance. Next, the distressed pavement images classified by experts were trained and tested for evaluation by three deep learning convolutional neural network (CNN) models: GoogLeNet, AlexNet, and VGGNet. The CNN models were image classification tools used to identify and extract the combined features of the target images via deep layers. Here, a data augmentation technique was applied to produce big distress data for training. Third, the dimensions of the detected distressed pavement patches were computed to estimate the quantity of repair materials needed. RESULTS : It was found that installing cameras 1.8 m above the ground on the exterior rear of the vehicle could provide clear pavement surface images with a resolution of 1 cm per pixel. The sensitivity analysis results of the trained GoogLeNet, AlexNet, and VGGNet models were 93 %, 86 %, and 72 %, respectively, compared to 62.7 % for the dimensional computation. Following readjustment of the image categories in the GoogLeNet model, distress detection sensitivity increased to 94.6 %. CONCLUSIONS : These findings support urgent maintenance by sending the detected distressed pavement images with the dimensions of the distressed patches and GPS coordinates to local maintenance offices in real-time.

4차 산업혁명 시대를 맞이하면서 첨단기술이 산업 전반에 사용됨에 따라 자율운항선박기술에 대한 관심이 고조되고 있다. 이에 본 연구에서는 자율운항선박에 대한 기초 연구로써 모형 선박을 제작하고 드론에 사용되는 아두파일럿(Ardupilot)을 적용시켜 자율운항 제어 시스템을 구축하였다. 구축된 모형선을 활용하여 다양한 항해계획을 실행함으로써 자율운항 선박의 운항 자동제어 가능성을 확인하였다. 침로안정성 실험에서는 모형선이 정해진 침로를 똑바로 따라가지 못하고 지그재그(S자 형태)로 항행하면서 침로에서 최대 5.4 m(4.5 L) 이탈하였으며, 매개변수를 수정하여 이탈거리를 최대 1.8 m(1.5 L)까지 감소시켰다. 선회성능 실험에서는 선회권의 직경이 최대 약 9.3 m(7.8 L) 로 나타났는데 매개변수를 수정하여도 큰 변화를 확인할 수 없었다. 하지만 WP 도착 전 감속하도록 실험한 결과 선회권의 직경이 최대 약 3.2 m(2.7 L)로 감소된 것을 확인할 수 있었다. 정지성능을 평가하기 위해 모든 실험의 마지막 정지 예정 위치와 실제 모형선이 정지한 위치를 상호 비교하였으며, 정지 예정 위치로부터 최소 0.4 m(0.3 L), 최대 6.2 m(5.2 L) 떨어진 지점에서 모형선이 정지하였음을 확인할 수 있었다. 추후 다양한 매개변수의 수정·보완을 통한 성능 향상이 이루어진다면 자동제어를 통한 자율운항선박의 자동접안을 위한 연구를 진행할 계획이다.

The purpose of this study was to develop an optimized hull-form of a coastal fishing vessel operating at high speed. In order to achieve the purpose of the study, HOTBOAT, that can perform an automatic hull-form optimal design of a coastal fishing vessel, was developed. HOTBOAT was composed of an objective function estimation algorithm and an optimization algorithm and a hull-form modification algorithm. In this study, the wave-making resistance was selected as an objective function and the potential-based panel method was applied to predict the objective function. SQP(sequential quadratic programming) method were adopted to predict the optimal direction and answer. Bell-shaped hull-form modification function method and NURBS(non-uniform rational B-spline) geometry modeling method were applied to modify the hull-form during the whole optimization process. HOTBOAT was applied to develop the optimal hull-form of a coastal fishing vessel with minimum wave resistance. The initial hull-form of the coastal fishing vessel was compared with the optimal hull-form. As a result of hull-form optimization, a coastal fishing vessel with a reduction of about 30% was developed compared with the initial hull-form and the displacement and the wetted surface area of the optimal hull-form was decreased to less than about 1% in comparison with the initial hull-form.