• 한국방재학회 논문집
• /
• 제8권4호
• /
• pp.23-29
• /
• 2008

본 연구는 사망률이 높은 도로 곡선부를 대상으로 도로설계요소를 기반으로 안전도 판단지수를 설정하고 이로부터 사고율을 산정하는 모듈과 곡선부와 곡선부 진입전 직선부에서의 속도차를 추정하는 모형을 개발하고 이로부터 곡선부의 안전도를 판단하는 등 두 개의 모듈을 제시하고 있다. 그리고 이러한 두 개의 모듈을 통합한 통합모델을 통해 곡선부의 안전도를 등급화 할 수 있도록 하는 것을 목적으로 한다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권3호
• /
• pp.861-880
• /
• 2023

Semantic segmentation of road scene is the key technology of autonomous driving, and the improvement of convolutional neural network architecture promotes the improvement of model segmentation performance. The existing convolutional neural network has the simplification of learning knowledge and the complexity of the model. To address this issue, we proposed a road scene semantic segmentation algorithm based on multi-task collaborative learning. Firstly, a depthwise separable convolution atrous spatial pyramid pooling is proposed to reduce model complexity. Secondly, a collaborative learning framework is proposed involved with saliency detection, and the joint loss function is defined using homoscedastic uncertainty to meet the new learning model. Experiments are conducted on the road and nature scenes datasets. The proposed method achieves 70.94% and 64.90% mIoU on Cityscapes and PASCAL VOC 2012 datasets, respectively. Qualitatively, Compared to methods with excellent performance, the method proposed in this paper has significant advantages in the segmentation of fine targets and boundaries.

• 한국수자원학회논문집
• /
• 제45권4호
• /
• pp.419-430
• /
• 2012

도로 배수시설 설계시 수리 및 수문요소 분석이 적절하게 적용되어야 하지만, 현재는 계산의 복잡함 때문에 충분히 고려되지 못하고 있다. 본 연구에서는 강우지속시간이 10분 이하인 도로배수유역에 적합한 분단위 강우강도식의 개발, 국내 도로배수유역의 지형 특성을 사실적으로 반영할 수 있는 운동파 모형 이론을 접목한 표면 박류 강우-유출 모형의 개발및검증, 노면배수시설설계, 암거단면규격산정및각종수로설계등의모형개발, 개발된 모형들을 도로설계자가 쉽게 익혀서 신속 정확하게 활용할 수 있도록 사용자 편의를 고려한 도로배수설계 전산프로그램을 개발하였다. 개발된 모형을 이용하여 적용성 검토를 수행하였으며, 현행 설계 방법과 개발된 설계 방법을 비교한 결과 노면 배수시설의 설치 간격은 6~65% 짧게 계산되었으며, 횡단 배수시설의 단면 크기는 6~140% 크게 계산되었다.

• 한국시스템다이내믹스연구
• /
• 제12권1호
• /
• pp.75-87
• /
• 2011

Congestion problems can be approached from the viewpoint of system dynamics theory. The relationship between road capacity and congestion can be explained by the 'relative control' archetype among four system archetypes suggested by Wolstenholme. There is a balancing feedback loop between road capacity and road congestion. However, there is another balancing loop between road congestion and car traffic volume, which keeps disrupting the equilibrium of the former loop. A system dynamics model, which is based on a partial adjustment model of induced traffic in the literature, is built to simulate three road building scenarios: 'Expanding investment', 'Balancing investment' and 'Frozen road investment' scenarios. The 'Expanding investment' scenario manages to drop congestion levels by 9% over 30 years, however, causing much higher emissions of $CO_2$ than other scenarios. The trade-off relationship between congestion levels and environmental costs must be taken into consideration for road investment policies.

• 한국소음진동공학회논문집
• /
• 제19권11호
• /
• pp.1144-1150
• /
• 2009

Appropriate vibration model is required to predict in advance the vibration level of a large vehicle which carries sensitive electronic/mechanical equipments and drives often on the unpaved and/or off-road conditions. Central tire inflation system(CTIS) is recently adopted to improve the mobile operation of military and bulletproof vehicles. In this paper, full vehicle model(FVM) having 11 degrees of freedom and equipped with CTIS has been developed for a large vehicle which has $8\times8$ wheels$\times$driving wheels. Usability of the developed model is validated via road tests for three different modes (i.e. highway, country, and mud/sand/snow modes) and for various velocity conditions. The developed FVM can be used to predict the vibration level of the large vehicle as well as to determine the driving velocity criterion for different road conditions.

• 한국안전학회지
• /
• 제25권6호
• /
• pp.186-196
• /
• 2010

It has been well known that traffic accidents occur under combined functional contributions of drivers, vehicles and road facilities, and that evaluation of safety levels for a specific road section or point is generally much complicated. Additionally, most of traffic accidents occur randomly implicating it is necessary to be evaluated in terms of probability theory. Thus, the evaluation model which reflects various characteristics and probabilistic distributions of traffic accidents has been necessary. The present paper provides a reliability based model with variables of probabilistic operating speeds and design speeds together which have been individually explaining associated characteristics in traffic accidents. Consequently, the model made it possible for speed management and road improvement projects to be evaluated in a common index. Application studies were performed in three cases. Through the studies, couples of facts were identified that the model successfully considered the probabilistic operating speeds and design speeds together and that then, the model evaluated road safety alternatives relatively which are complicatedly characterized and differently located.

• 한국환경과학회지
• /
• 제13권6호
• /
• pp.505-512
• /
• 2004

This study was attempted to review the information data for minimizing the destruction of environmental naturalness and the visual damage of landscape from road construction by establishing a model of environmental naturalness for national roads in the suburb areas to suggest an answer to a research question, ' hat does decide the environmental naturalness of roadscape?'. We found that 1) The road-side slope showed no statistical significance in the description of environmental naturalness of roadscape, but the fact that the road-side slope from road construction is the destruction of natural topography cannot be overlooked. 2) In terms of the direction of value variations for independent variables, signboard and telegraph post, soundproofing and protection wall, structure, and building acted toward negative (-) direction, while mountains, sky, road trees, fields, and surrounding green including the road-side slope acted toward positive(+) direction. 3) The variable with highest relative contribution to dependent variables among independent variables is building, which has importance as many as 148 times of road-side slope, while the variable road-side slope has the least importance. Building has the importance of 7.22 times, mountains 5.51 times, road trees 2.59 times, surrounding green 2.54 times, structure 2.41 times, signboard and telegraph post 2.37 times, soundproofing and protection wall 2.20 times, and sky 1.32 times of the fields as a standard criterion values 1.

• 한국소음진동공학회논문집
• /
• 제19권4호
• /
• pp.425-432
• /
• 2009

Recently, reduction of road traffic noise in residential buildings has become one of the most important subjects. To reduce the road traffic noise, noise impact assessment by the road traffic prediction model is required before building construction. For reasonable road traffic noise prediction, it is required to analysis of various factors in road traffic prediction models. This paper was studied the road traffic noise propagation factors such as distance from road to building, receiver height, alignment angle of building and reflection coefficient of the building facade by two calculation models, RLS-90 and CRTN. The result showed that noise reduction was generally higher at bottom stories by ground absorption effect. The reflection coefficient of the building facade was affect of additional sound pressure level by facade reflecting. And alignment angle of building at $90^{\circ}$ was performed effective noise reduction better than $0^{\circ}$.

• 환경영향평가
• /
• 제14권6호
• /
• pp.427-441
• /
• 2005

In order to choose proper dispersion model and emission factors suitable in Korea in evaluating the effect of pollutants emitted by the vehicles in highway on nearby area, various road dispersion models and vehicle emission factors were reviewed. With theoretical inter-comparisons of the exiting models for line source, CALINE 3 and CALINE 4 models which were suggested by US EPA were selected as the road dispersion models for further evaluation with the measurement. The emission factors suggested by Korean Ministry of Environment was turned out to be appropriate since the classification of vehicle kinds was simple and easy to apply in Korea. The comparisons of predicted concentrations by CALINE 3 and 4 models with the measurements in flat, fill and bridge road types showed that CO and PM-10 were in good agreements with experiments and the differences between CALINE 3 and 4 models are negligible. The model concentrations of $NO_2$ by CALINE 4 were also in good agreement with the measurement but those by CALINE 3 were over-predicted. The discrepancies in CALINE 3 model were due to rapid decay reaction of $NO_2$ near the highway, which was not included in CALINE 3 model. For the road type with one & two side cutting grounds, the similar patterns as the flat & fill road type for CO, PM10, & $NO_2$ were observed but the number of data for comparison in these cases were not enough to draw the conclusion. These results lead to the conclusion that CALINE4 model is proper in road environmental impact assessment near the highway in flat, fill and bridge road types.

• 대한원격탐사학회지
• /
• 제33권4호
• /
• pp.423-436
• /
• 2017

This paper presents a deep learning-based road segmentation framework from very high-resolution orthophotos. The proposed method uses Deep Convolutional Autoencoders for end-to-end mapping of orthophotos to road segmentations. In addition, a set of post-processing steps were applied to make the model outputs GIS-ready data that could be useful for various applications. The optimization of the model's parameters is explained which was conducted via grid search method. The model was trained and implemented in Keras, a high-level deep learning framework run on top of Tensorflow. The results show that the proposed model with the best-obtained hyperparameters could segment road objects from orthophotos at an average accuracy of 88.5%. The results of optimization revealed that the best optimization algorithm and activation function for the studied task are Stochastic Gradient Descent (SGD) and Exponential Linear Unit (ELU), respectively. In addition, the best numbers of convolutional filters were found to be 8 for the first and second layers and 128 for the third and fourth layers of the proposed network architecture. Moreover, the analysis on the time complexity of the model showed that the model could be trained in 4 hours and 50 minutes on 1024 high-resolution images of size $106{\times}106pixels$, and segment road objects from similar size and resolution images in around 14 minutes. The results show that the deep learning models such as Convolutional Autoencoders could be a best alternative to traditional machine learning models for road segmentation from aerial photographs.