• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.16 no.4
• /
• pp.153-163
• /
• 2017

This research is based on the previous research that personally preferred safe distance, rotating angle and speed are differentiated. Thus, we use machine learning model for recognizing maneuvering modes trained per personal or per similar driving pattern groups, and we evaluate automatic driving according to maneuvering modes. By utilizing driving knowledge, we subdivided 8 kinds of longitudinal modes and 4 kinds of lateral modes, and by combining the longitudinal and lateral modes, we build 21 kinds of maneuvering modes. we train the labeled data set per time stamp through RNN, LSTM and Bi-LSTM models by the trips of drivers, which are supervised deep learning models, and evaluate the maneuvering modes of automatic driving for the test data set. The evaluation dataset is aggregated of living trips of 3,000 populations by VTTI in USA for 3 years and we use 1500 trips of 22 people and training, validation and test dataset ratio is 80%, 10% and 10%, respectively. For recognizing longitudinal 8 kinds of maneuvering modes, RNN achieves better accuracy compared to LSTM, Bi-LSTM. However, Bi-LSTM improves the accuracy in recognizing 21 kinds of longitudinal and lateral maneuvering modes in comparison with RNN and LSTM as 1.54% and 0.47%, respectively.

• Journal of Korean Society of Transportation
• /
• v.22 no.6
• /
• pp.89-99
• /
• 2004

The purpose of this study is to evaluate the value of information on Bus-Route concerning on the User's Individual value. The value of information is estimated with the price of time saving by using the information. The price of unit time for each user is applied to convert the saving time to the cost. To estimate the user's expense from origin to destination the previous model is modified. Bus-travel cost is estimated with variables such as bus-travel time, bus-interval, bus-fare, and the price of walking distance. In this study, to estimate in-vehicle time the bus-travel time model is developed based on the spatial characteristics distinguished by three types of circular-road in the network of Daegu Metropolitan area. For the case study, a set of the origin and destination is selected as Dalsu-gu District Office and East Daegu Train Station respectively. There are several bus-routes which can be used as direct or transferable bus-routes selected. The study showed that when the value of time for individual users is \1,738/hr, there is no benefit to using information of transferable bus-routes. It also showed that the more discount rates of bus fare is increased, the benefit to using information of transferable bus-routes is increased, and that the lower value of time is, the benefit to using information of transferable bus-routes is increased.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.41 no.1
• /
• pp.75-83
• /
• 2021

Over-urbanization has contributed to the increase in traffic problems. This makes the need for effective road planning and design more important than ever. I have been able to learn how to build a new road, and how to use it. However, in spite of the importance of good road planning, there are no systematic standards or methods for calculating traffic volume on railroad routes. Therefore, in this study, to strengthen the competitiveness of railroads, the concept of line capacity is introduced to railroads, and a clear standard and method for calculating railroad line capacity are presented. Based on the results, the line capacity of main railway lines for domestic railways was calculated. By applying the method of calculating the line capacity presented in this study, the capacity of existing railway lines and newly expanded routes can be calculated. It is expected that our findings will be able to provide systematic standards that can be applied to yield a more effective investment and design planning stage; the findings will also help improve the efficiency of railroad operation.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.1-16
• /
• 2022

Despite recent breakthroughs in deep learning and computer vision fields, the pixel-wise identification of tiny objects in high-resolution images with complex disturbances remains challenging. This study proposes a modified U-net for tiny crack segmentation in real-world steel-box-girder bridges. The modified U-net adopts the common U-net framework and a novel Self-Attention-Self-Adaption (SASA) neuron as the fundamental computing element. The Self-Attention module applies softmax and gate operations to obtain the attention vector. It enables the neuron to focus on the most significant receptive fields when processing large-scale feature maps. The Self-Adaption module consists of a multiplayer perceptron subnet and achieves deeper feature extraction inside a single neuron. For data augmentation, a grid-based crack random elastic deformation (CRED) algorithm is designed to enrich the diversities and irregular shapes of distributed cracks. Grid-based uniform control nodes are first set on both input images and binary labels, random offsets are then employed on these control nodes, and bilinear interpolation is performed for the rest pixels. The proposed SASA neuron and CRED algorithm are simultaneously deployed to train the modified U-net. 200 raw images with a high resolution of 4928 × 3264 are collected, 160 for training and the rest 40 for the test. 512 × 512 patches are generated from the original images by a sliding window with an overlap of 256 as inputs. Results show that the average IoU between the recognized and ground-truth cracks reaches 0.409, which is 29.8% higher than the regular U-net. A five-fold cross-validation study is performed to verify that the proposed method is robust to different training and test images. Ablation experiments further demonstrate the effectiveness of the proposed SASA neuron and CRED algorithm. Promotions of the average IoU individually utilizing the SASA and CRED module add up to the final promotion of the full model, indicating that the SASA and CRED modules contribute to the different stages of model and data in the training process.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.26 no.3
• /
• pp.37-51
• /
• 2023

Spatial information is essential for interpreting urban phenomena. Methodologies for spatializing urban information, especially when it lacks location details, have been consistently developed. Typical methods include Geocoding using structured address information or place names, spatial integration with existing geospatial data, and manual tasks utilizing reference data. However, a vast number of documents produced by administrative agencies have not been deeply dealt with due to their unstructured nature, even when there's demand for spatialization. This research utilizes the natural language processing model BERT to spatialize public documents related to urban planning. It focuses on extracting sentence elements containing addresses from documents and converting them into structured data. The study used 18 years of urban planning public announcement documents as training data to train the BERT model and enhanced its performance by manually adjusting its hyperparameters. After training, the test results showed accuracy rates of 96.6% for classifying urban planning facilities, 98.5% for address recognition, and 93.1% for address cleaning. When mapping the result data on GIS, it was possible to effectively display the change history related to specific urban planning facilities. This research provides a deep understanding of the spatial context of urban planning documents, and it is hoped that through this, stakeholders can make more effective decisions.

• Steel and Composite Structures
• /
• v.49 no.1
• /
• pp.81-89
• /
• 2023

A simplified calculation method of natural vibration characteristics of high-speed railway multi-span bridge-longitudinal ballastless track system is proposed. The rail, track slab, base slab, main beam, bearing, pier, cap and pile foundation are taken into account, and the multi-span longitudinal ballastless track-beam-bearing-pier-cap-pile foundation integrated model (MBTIM) is established. The energy equation of each component of the MBTIM based on Timoshenko beam theory is constructed. Using the improved Fourier series, and the Rayleigh-Ritz method and Hamilton principle are combined to obtain the extremum of the total energy function. The simplified calculation formula of the natural vibration frequency of the MBTIM under the influence of vertical and longitudinal vibration is derived and verified by numerical methods. The influence law of the natural vibration frequency of the MBTIM is analyzed considering and not considering the participation of each component of the MBTIM, the damage of the track interlayer component and the stiffness change of each layer component. The results show that the error between the calculation results of the formula and the numerical method in this paper is less than 3%, which verifies the correctness of the method in this paper. The high-order frequency of the MBTIM is significantly affected considering the track, bridge pier, pile soil and pile cap, while considering the influence of pile cap on the low-order and high-order frequency of the MBTIM is large. The influence of component damage such as void beneath slab, mortar debonding and fastener failure on each order frequency of the MBTIM is basically the same, and the influence of component damage less than 10m on the first fourteen order frequency of the MBTIM is small. The bending stiffness of track slab and rail has no obvious influence on the natural frequency of the MBTIM, and the bending stiffness of main beam has influence on the natural frequency of the MBTIM. The bending stiffness of pier and base slab only has obvious influence on the high-order frequency of the MBTIM. The natural vibration characteristics of the MBTIM play an important guiding role in the safety analysis of high-speed train running, the damage detection of track-bridge structure and the seismic design of railway bridge.

• Geophysics and Geophysical Exploration
• /
• v.26 no.4
• /
• pp.171-184
• /
• 2023

Machine learning (ML)-based cavity detection using a large amount of survey data obtained from vehicle-mounted ground penetrating radar (GPR) has been actively studied to identify underground cavities. However, only simple image processing techniques have been used for preprocessing the ML input, and many conventional seismic and GPR data processing techniques, which have been used for decades, have not been fully exploited. In this study, based on the idea that a cavity can be identified using diffraction, we applied ML-based diffraction separation to GPR data to increase the accuracy of cavity detection using the YOLO v5 model. The original ML-based seismic diffraction separation technique was modified, and the separated diffraction image was used as the input to train the cavity detection model. The performance of the proposed method was verified using public GPR data released by the Seoul Metropolitan Government. Underground cavities and objects were more accurately detected using separated diffraction images. In the future, the proposed method can be useful in various fields in which GPR surveys are used.

• Journal of the Korean Geotechnical Society
• /
• v.39 no.12
• /
• pp.37-46
• /
• 2023

Dynamic soil properties are essential factors for predicting the detailed behavior of the ground. However, there are limitations to gathering soil samples and performing additional experiments. In this study, we used an artificial neural network (ANN) to predict dynamic soil properties based on static soil properties. The selected static soil properties were soil cohesion, internal friction angle, porosity, specific gravity, and uniaxial compressive strength, whereas the compressional and shear wave velocities were determined for the dynamic soil properties. The Levenberg-Marquardt and Bayesian regularization methods were used to enhance the reliability of the ANN results, and the reliability associated with each optimization method was compared. The accuracy of the ANN model was represented by the coefficient of determination, which was greater than 0.9 in the training and testing phases, indicating that the proposed ANN model exhibits high reliability. Further, the reliability of the output values was verified with new input data, and the results showed high accuracy.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.16 no.6
• /
• pp.385-391
• /
• 2023

As the use of electric vehicles has increased to minimize carbon emissions, the analyzing the state and performance of lithium-ion batteries that is instrumental in electric vehicles have been important. Comprehensive analysis using not only the voltage, current and temperature of the battery pack, which can affect the condition and performance of the battery, but also the driving data and charging pattern data of the electric vehicle is required. Therefore, a thorough analysis is imperative, utilizing electric vehicle operation data, charging pattern data, as well as battery pack voltage, current, and temperature data, which collectively influence the condition and performance of the battery. Therefore, collection and preprocessing of battery data collected from electric vehicles, collection and preprocessing of data on driver driving habits in addition to simple battery data, detailed design and modification of artificial intelligence algorithm based on the analyzed influencing factors, and A battery analysis and evaluation model was designed. In this paper, we gathered operational data and battery data from real-time electric buses. These data sets were then utilized to train a Random Forest algorithm. Furthermore, a comprehensive assessment of battery status, operation, and charging patterns was conducted using the explainable Artificial Intelligence (XAI) algorithm. The study identified crucial influencing factors on battery status, including rapid acceleration, rapid deceleration, sudden stops in driving patterns, the number of drives per day in the charging and discharging pattern, daily accumulated Depth of Discharge (DOD), cell voltage differences during discharge, maximum cell temperature, and minimum cell temperature. These factors were confirmed to significantly impact the battery condition. Based on the identified influencing factors, a battery analysis and evaluation model was designed and assessed using the Random Forest algorithm. The results contribute to the understanding of battery health and lay the foundation for effective battery management in electric vehicles.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.199-207
• /
• 2024

Local reservoirs are crucial sources for agricultural water supply, necessitating stable water level management to prepare for extreme climate conditions such as droughts. Water level prediction is significantly influenced by local climate characteristics, such as localized rainfall, as well as seasonal factors including cropping times, making it essential to understand the correlation between input and output data as much as selecting an appropriate prediction model. In this study, extensive multivariate data from over 400 reservoirs in Jeollabuk-do from 1991 to 2022 was utilized to train and validate a water level prediction model that comprehensively reflects the complex hydrological and climatological environmental factors of each reservoir, and to analyze the impact of each input feature on the prediction performance of water levels. Instead of focusing on improvements in water level performance through neural network structures, the study adopts a basic Feedforward Neural Network composed of fully connected layers, batch normalization, dropout, and activation functions, focusing on the correlation between multivariate input data and prediction performance. Additionally, most existing studies only present short-term prediction performance on a daily basis, which is not suitable for practical environments that require medium to long-term predictions, such as 10 days or a month. Therefore, this study measured the water level prediction performance up to one month ahead through a recursive method that uses daily prediction values as the next input. The experiment identified performance changes according to the prediction period and analyzed the impact of each input feature on the overall performance based on an Ablation study.