• Journal of Korean Society of Transportation
• /
• v.28 no.1
• /
• pp.125-134
• /
• 2010

It is occurred bus arrival time errors when a bus arrives at a bus stop because of a variety of traffic condition such as traffic signal cycle, the time to get on and off a bus, a bus-only lane and so on. In this paper, bus delay time which is occurred as the result of traffic condition was estimated with Markov Chain process and bus arrival time at each bus stop was predicted with it. As the result of the study, it is confirmed to improve accuracy than the method of bus arrival time prediction with existing method (weighed moving average method) in case predicting bus arrival time using 7 by 7 and 9 by 9 matrixes.

• Proceedings of the KAIS Fall Conference
• /
• 2009.05a
• /
• pp.672-675
• /
• 2009

BMS(Bus Management System)의 핵심인 버스도착예정시간을 산출하는 데 있어서 기존 대부분의 도시에서는 시계열 모형의 이동평균법, 칼만필터링 등으로 버스도착예정시간을 예측하고 있으나 이는 급격한 통행량의 변화 또는 급작스러운 사고, 신호체계 등에 적응 할 수 없다. 따라서 본 논문에서는 주변 도로의 통행량에 따른 버스의 정류장 도착시간을 예측하는 방법을 제안 한다. 주변 도로의 통행량과 실제 버스의 통행시간을 실측하여 기록, 학습하고 모델링하여 미래의 버스의 운행시간을 예측하는 방법이다. 또, 이동평균법에 의한 버스도착시간 예측결과와 본 논문에서 제안하는 결과와 비교, 분석하였다.

• KIPS Transactions on Computer and Communication Systems
• /
• v.6 no.4
• /
• pp.189-196
• /
• 2017

BIS(Bus Information System) provides the different information related to buses including predictions of arriving times at stations. BIS have been deployed almost all cities in our country and played active roles to improve the convenience of public transportation systems. Moving average filters, Kalman filter and regression models have been representative in forecasting the arriving times of buses in current BIS. The accuracy in prediction of arriving times depends largely on the forecasting algorithms and traffic conditions considered when forecasting in BIS. In present BIS, the simple prediction algorithms are used only considering the passage times and distances between stations. The forecasting of arrivals, however, have been influenced by the traffic conditions such as traffic signals, traffic accidents and pedestrians ets., and missing data. To improve the accuracy of bus arriving estimates, there are big troubles in building models including the above problems. Hidden Markov Models have been effective algorithms considering various restrictions above. So, we have built the HMM forecasting models for bus arriving times in the current BIS. When building models, the data collected from Sunchean City at 2015 have been utilized. There are about 2298 stations and 217 routes in Suncheon city. The models are developed differently week days and weekend. And then the models are conformed with the data from different districts and times. We find that our HMM models can provide more accurate forecasting than other existing methods like moving average filters, Kalmam filters, or regression models. In this paper, we propose Hidden Markov Model to obtain more precise and accurate model better than Moving Average Filter, Kalman Filter and regression model. With the help of Hidden Markov Model, two different sections were used to find the pattern and verified using Bootstrap process.

• 한국ITS학회:학술대회논문집
• /
• 2005.11a
• /
• pp.40-45
• /
• 2005

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.8 no.3
• /
• pp.1-10
• /
• 2009

Bus delay time is occurred as the result of traffic condition and important factor to predict bus arrival time. In this paper, transition probability matrixes between bus stops are made by using Markov Chain and it is predicted bus delay time with them. As the results of study, it is confirmed a possibility of adapting the assumption which it has same bus transition probability between stops through paired-samples T-test and overcame the limitation of exiting studies in case there is no scheduled bus arrival time for each stops with using bus interval time. Therefore it will be possible to predict bus arrival time with Markov Chain.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.16 no.6
• /
• pp.101-111
• /
• 2017

This study propose a method to predict the bus arrival time by considering the signal delay time which is an element which can not be considered in the current bus arrival prediction information generation algorithm. In order to consider the signal delay time, travel time is divided into three components: service time, cruising travel time, and signal delay time. Signal delay time was estimated using intersection arrival time and TOD. The results show that most of the errors that occurred in predicting the arrival time are within about 30 seconds. Some of the estimates have large errors due to the nature of this methodology that uses the estimated value of the intersection arrival time rather than the observation value. It is also difficult to predict the arrival time of the express buses using this method. Future studies such as improving this through real-time location information will greatly improve the accuracy of the methodology.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.39 no.6
• /
• pp.833-839
• /
• 2019

Bus Information System (BIS) collects information related to the operation of buses and provides information to users through predictive algorithms. Method of predicting through recent information in same section reflects the traffic situation of the section, but cannot reflect the characteristics of the target line. The method of predicting the historical data at the same time zone is limited in forecasting peak time with high volatility of traffic flow. Therefore, we developed a pattern recognition bus arrival time prediction algorithm which could be overcome previous limitation. This method recognize the traffic pattern of target flow and select the most similar past traffic pattern. The results of this study were compared with the BIS arrival forecast information history of Seoul. RMSE of travel time between estimated and observed was approximately 35 seconds (40 seconds in BIS) at the off-peak time and 40 seconds (60 seconds in BIS) at the peak time. This means that there is data that can represent the current traffic situation in other time zones except for the same past time zone.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.338-339
• /
• 2021

본 논문에서는 제주시에서 오픈데이터로 공개한 버스탑승 기록을 기반으로 이용도가 높은 버스노선에 대해 특정정류장에서의 도착시간 예측모델을 구축한다. 버스들의 평균주행 속도, 운행시간대, 교통량 등을 입력으로 한 모델을 Sklearn을 이용하여 생성하고 MAE와 손실율 등의 성능을 분석한다.

• Journal of Korean Society of Transportation
• /
• v.20 no.1
• /
• pp.7-18
• /
• 2002

To activate public transportation service, Bucheon City built Bus Information System based on Beacon type, and operates it for no.22 line. This research analyzes an effect of BIS operations, and mainly it analyzes far reliability evaluation of bus arrival time information and passenger satisfaction about BIS. As results of reliability evaluation of arrival time information service, it is proven to be practically inappropriate to use as arrival time data because it is not only travel time between each bus stop but also previous travel time history data. In order to improve this matter, neural network model was evaluated as the most outstanding one as result of experiment in applying current arrival time Prediction model. This research cannot help limiting for evaluation of operation effect in Bucheon City because there is no Bus Information System based on GPS type in Korea. For the future ITS model city, in the case of building ITS model city based on GPS type, it is possible to compare two systems relatively. In addition to that, fur the consideration of reliability of bus arrival time information, it is required to develop Predictable model and research factors that affect to bus operation.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.4 no.3 s.8
• /
• pp.51-59
• /
• 2005

To improve the quality of bus service, providing bus ravel time information to passenger through station screen. Generally, bus travel time information predict by using previous bus data such as neural network, Kalman filtering, and moving average algorithms. However, when they got a difficulty about bus travel time information because of the missing previous bus data, they use pattern data. Generally, nevertheless the difference of range is big. Hence in this research to calculate the bus travel time information when the bus information is missed, use queue detector's data which set up in link. The application of several factors which influence in bus link travel time, we used CORSIM Version 5.1 simulation package.