• Journal of Korean Society of Transportation
• /
• v.32 no.3
• /
• pp.266-279
• /
• 2014

For the evaluation of roadway safety, diverse methods, including before-after studies, simple comparison using historic traffic accident data, methods based on experts' opinion or literature, have been applied. Especially, many research efforts have developed traffic accident prediction models in order to identify critical elements causing accidents and evaluate the level of safety. A traffic accident prediction model must secure predictability and transferability. By acquiring the predictability, the model can increase the accuracy in predicting the frequency of accidents qualitatively and quantitatively. By guaranteeing the transferability, the model can be used for other locations with acceptable accuracy. To this end, traffic accident prediction models using non-linear regression, artificial neural network, and structural equation were developed in this study. The predictability and transferability of three models were compared using a model development data set collected from 90 signalized intersections and a model validation data set from other 33 signalized intersections based on mean absolute deviation and mean squared prediction error. As a result of the comparison using the model development data set, the artificial neural network showed the highest predictability. However, the non-linear regression model was found out to be most appropriate in the comparison using the model validation data set. Conclusively, the artificial neural network has a strong ability in representing the relationship between the frequency of traffic accidents and traffic and road design elements. However, the predictability of the artificial neural network significantly decreased when the artificial neural network was applied to a new data which was not used in the model developing.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.15 no.6
• /
• pp.1123-1130
• /
• 2020

This study aims to classify the severity in car crashes using five classification learning models. The dataset used in this study contains 21,013 vehicle crashes, obtained from Korea Expressway Corporation, between the year of 2015-2017 and the LightGBM(Light Gradient Boosting Model) performed well with the highest accuracy. LightGBM, the number of involved vehicles, type of accident, incident location, incident lane type, types of accidents, types of vehicles involved in accidents were shown as priority factors. Based on the results of this model, the establishment of a management strategy for response of highway traffic accident should be presented through a consistent prediction process of accident severity level. This study identifies applicability of Machine Learning Models for Predicting of the Severity of Car Traffic Accidents on a Highway and suggests that various machine learning techniques based on big data that can be used in the future.

• International Journal of Highway Engineering
• /
• v.15 no.6
• /
• pp.113-123
• /
• 2013

PURPOSES : Although numerous researches have been studied to reveal accident causations for road intersections, there are still many research gaps for road segments. It is mainly because of difficulty of data and lack of analytical method. This study aims to study accident causations for rural road segments and develop accident modification factors for safety evaluation. The accident modification factors can be used to improve road safety. METHODS : Methods for developing AMF are diverse. This study developed AMFs using accident prediction models and selected explanatory variables from the accident models. In order to select final AMFs, three different methods were applied in the study. RESULTS : As a result of the study, many AMFs such as horizontal curves or vertical curves were developed and explained the meanings of the results. CONCLUSIONS : This study introduced meaningful methods for developing significant AMFs and also showed several AMFs. It is expected that traffic or road engineers will be able to use the AMFs to improve road segment safety.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.15 no.4
• /
• pp.68-80
• /
• 2016

This study proposes the traffic prediction and optimal traffic control system based on cell transmission model and genetic algorithm in cloud environment. The proposed prediction and control system consists of four parts. 1) Data preprocessing module detects and imputes the corrupted data and missing data points. 2) Data-driven traffic prediction module predicts the future traffic state using Multi-level K-Nearest Neighbor (MK-NN) Algorithm with stored historical data in SQL database. 3) Online traffic simulation module simulates the future traffic state in various situations including accident, road work, and extreme weather condition with predicted traffic data by MK-NN. 4) Optimal road control module produces the control strategy for large road network with cell transmission model and genetic algorithm. The results show that proposed system can effectively reduce the Vehicle Hours Traveled upto 60%.

• Journal of Korean Society of Transportation
• /
• v.34 no.4
• /
• pp.341-353
• /
• 2016

Highway is mainly in charge of middle-long distance of vehicular travel. Trip length has shown a growing trend due to increased commute distances by the relocation of public agencies. For this reason, the proportion of driver-driven accidents, caused by their fatigue or sleepiness, are very high on highways. However, existing studies related to accident prediction have mainly considered external factors, such as road conditions, environmental factors and vehicle factors, without driving behavior. In this study, we suggested an accident index (FDR, Fatigued Driving Rate) based on traffic behavior using large-scale Car Navigation path data, and exlpored the relationship between FDR and traffic accidents. As a result, FDR and traffic accidents showed a high correlation. This confirmed the need for a paradigm shift (from facilities to travel behavior) in traffic accident prediction studies. FDR proposed in this study will be utilized in a variety of fields. For example, in providing information to prevent traffic accidents (sleepiness, reckless driving, etc) in advance, utilization of core technologies in highway safety diagnostics, selection of priority location of rest areas and shelter, and selection of attraction methods (rumble strips, grooving) for attention for fatigued sections.

• International Journal of Highway Engineering
• /
• v.13 no.1
• /
• pp.69-77
• /
• 2011

This study is to suggest a methodology to overcome the uncertainty and lack of reliability of data. The fuzzy reasoning model and the neural network model were developed in order to overcome the potential lack of reliability which may occur during the process of data collection. According to the result of comparison with the Poisson regression model, the suggested models showed better performance in the accuracy of the accident frequency prediction. It means that the more accurate accident frequency prediction model can be developed by the process of the uncertainty of raw data and the adjustment of errors in data by learning. Among the suggested models, the performance of the neural network model was better than that of the fuzzy reasoning model. The suggested models can evaluate the safety of signalized intersections in operation and/or planning, and ultimately contribute the reduction of accidents.

• Journal of Korean Society of Transportation
• /
• v.30 no.6
• /
• pp.37-46
• /
• 2012

The number of traffic accidents caused by elderly drivers over the age of 65 has surged over the past ten years from 37,000 to 274,000 cases. The proportion of elderly drivers' accidents has jumped 3.1 times from 1.2% to 3.7% out of all traffic accidents, and traffic safety organizations are pursuing diverse measures to address the situation. Above all, connecting safety measures with an in-depth research on behavioral and physical characteristics of elderly drivers will prove vital. This study conducted an empirical research linking the driving characteristics and traffic accidents by elderly drivers based on the Driving Aptitude Test items and traffic accident data, which enabled the measurement of behavioral characteristics of elderly drivers. In developing the Influence Model, we applied the zero-inflated Poisson (ZIP) regression model and selected an accident prediction model based on the Bayesian Influence in regards to the ZIP regression model and the zero-inflated negative binomial (ZINB) regression model. According to the results of the AAE analysis, the ZIP regression model was more appropriate and it was found that three variables? prediction of velocity, diversion, and cognitive ability? had a relation of influence with traffic accidents caused by elderly drivers.

• Journal of Navigation and Port Research
• /
• v.33 no.8
• /
• pp.519-524
• /
• 2009

Although it is expected to provide fundamental data for advanced VTS system by analyzing traffic accidents in VTS area, there is no quantitative analysis to find it.. In this research, it is examined and analyzed marine casualties records(1999-2004), data of Port-MIS and data of each VTS center. The results of this research are as below. 1) It is necessary to reduce traffic accident and to improve VTS operating system. 2) It is discovered for statistical discrepancy between vessels controlled by VTS and vessels not controlled by VTS in accident cause, visibility, perception distance and cause of late perception in collision accidents 3) It is necessary for VTS assistance to be positive and to made in ample time consecutively. 4) As the result of traffic accident prediction model, it is necessary to develop a system improving VTS operators' ability to identify dangerous ships.

• Journal of Korean Society of Transportation
• /
• v.23 no.2
• /
• pp.95-106
• /
• 2005

Road accidents are considered as the result of a complex interplay between road, vehicle, environments, and human factors. Little study, however, has been carried out on the attributes of human factor compared to the road geometric conditions and traffic conditions. The previous researches focused on mainly both traffic and geometric conditions on specific location. Therefore, it's hard to explain phenomenon of the high traffic accident rates where road and traffic conditions are good. Because of these reasons, accident analysis has contributed on geometric improvement and has not contributed on traffic management such as selection of attention section, driver napping alert, etc. The freeway incident management is also associated with reliable prediction of incident occurrences on freeway sections. This paper presents a method for estimating the effect of trip length on freeway accident rate. A PAR (Potential Accident Ratio), the new concept of accident analysis, considering TLFDs (Trip Length Frequency Distributions) is suggested in this paper. This approach can help to strengthen freeway management and to reduce the likelihood of accidents.

• Proceedings of the KOR-KST Conference
• /
• 1997.12a
• /
• pp.33-69
• /
• 1997

This study was conducted to develop a methodology to predict utility pole accident rates and to evaluate cost-effectiveness for safety improvement for utility pole accidents. The utility pole accident rate prediction model was based on the encroachment rate approach introduced in the Transportation Research Board Special Report 214. The utility pole accident rate on a section of highway depends on the roadside encroachment rate and the lateral extent of encroachment. The encroachment rate is influenced by the horizontal and vertical alignment of the highway as well as traffic volume and mean speed. The lateral extent of encroachment is affected by the horizontal and vertical alignment, the mean speed and the roadside slope. An analytical method to generate the probability distribution function for the lateral extent of encroachment was developed for six kinds of encroachment types by the horizontal alignment and encroachment direction. The encroachment rate was calibrated with the information on highway and roadside conditions and the utility pole accident records collected on the sections of 55mph speed limit of the State Trunk Highway 12 in Wisconsin. The encroachment rate on a tangent segment was calibrated as a function of traffic volume with the actual average utility pole accident rates by traffic volume strategies. The adjustment factors for horizontal and vertical alignment were then derived by comparing the actual average utility pole accident rates to the estimations from the model calibrated for tangent and level sections. A computerized benefit-cost analysis procedure was then developed as a means of evaluating alternative countermeasures. The program calculates the benefit-cost ratio and the percent of reduction of utility pole accidents resulting from the implementation of a safety improvement. This program can be used to develop safety improvement: alternatives for utility pole accidents when a predetermined performance level is specified.