• The Korean Journal of Applied Statistics
• /
• v.16 no.2
• /
• pp.203-215
• /
• 2003

In order to provide the drivers with more reliable transportation information in NHTMS(National Highway Transportation Management System), it is important to estimate the expected passage time by using the traffic volume and speed. In this study, we analyze the characteristics of the traffic volume in the national highway and we investigate two real-time adjustment methods: the average adjustment method and the auto-regressive adjustment method. In addition, we compare them using the real data collected at the National Highway Route 3 in 2000.

• KIPS Transactions on Computer and Communication Systems
• /
• v.12 no.2
• /
• pp.53-60
• /
• 2023

Recently, the demand and traffic volume for various multimedia contents are rapidly increasing through real-time streaming platforms. In this paper, we predict real-time streaming traffic to improve the quality of service (QoS). Statistical models have been used to predict network traffic. However, since real-time streaming traffic changes dynamically, we used recurrent neural network-based deep learning models rather than a statistical model. Therefore, after the collection and preprocessing for real-time streaming data, we exploit vanilla RNN, LSTM, GRU, Bi-LSTM, and Bi-GRU models to predict real-time streaming traffic. In evaluation, the training time and accuracy of each model are measured and compared.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.10a
• /
• pp.362-364
• /
• 2018

Predicting network traffic volume has become a popular topic recently due to its support in many situations such as detecting abnormal network activities and provisioning network services. Especially, predicting the volume of the next upcoming traffic from the series of observed recent traffic volume is an interesting and challenging problem. In past, various techniques are researched by using time series forecasting methods such as moving averaging and exponential smoothing. In this paper, we propose a long short-term memory neural network (LSTM) based network traffic volume prediction method. The proposed method employs the changing rate of observed traffic volume, the corresponding time window index, and a seasonality factor indicating the changing trend as input features, and predicts the upcoming network traffic. The experiment results with real datasets proves that our proposed method works better than other time series forecasting methods in predicting upcoming network traffic.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.11 no.1
• /
• pp.74-81
• /
• 1998

It is not good joint that today's traffic control system that the course of traffic volume increase tendency is followed, in the traffic volume is approched into the time of my car. Accordingly when we analyzed the existing traffic signal control system, the traffic signal system is developed from the machine type that the motor was centered, to get up to date, to the intelligence electron signal control system. But yet, when we have a test and a A/S on the control circuit, the circuit that is designed to the center IC and ROM are complicated. Also, the time of pass lamp that the car line stream is going, can not extended automatically a time till the traffic volume is decreased to the same direction. This theme must be a real time intelligence control system that the time of pass lamp can extend aumatically. The circuit of sequence ladder diagram on the traffic signal control of a crossroads that is desinged, can be satisfied the complicated vehicle order. Therefore when the circuit is changed, the new developed system is economical with that dosen't needs any of components to require the circuit equipment, and the time is saved with needlessness of the circuit wiring again, and have a much trustworthy. The control method of pass signal lamp in the car line stream connecting among PLC and Relay and Temp Sensor, can be changed to hand operation and to semi-automation and to all-automation. New intelligence traffic signal system is composed with all-together system of T Sensor + Video Camera + IBM PC that is able to guiding the establishment of traffic order.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.9 no.3
• /
• pp.14-24
• /
• 2006

There are many problems and limits in equipments being used for traffic-volume analysis in the country. And traffic-volume information acquired through existing equipments is not provided in real-time. In the case of urban, there are limits on guarantee of trust on comprehending a appropriate road-volume because of difficulty on analyzing traffic-volume density and time series. And it is difficult to applicate in deciding a road policy as existing equipments don't provide the control information of traffic-flow. Therefore, it is necessary to build a road-flow policy rapidly and accurately through the road-flow information that analyze post-processed statistics data using traffic-flow investigation based on real time. In this study, we developed TICS(Traffic Information Collection System) based on GPS which could transmit traffic information transformed from car location information to traffic control center. And we developed TCS(Traffic Control System) based on Web GIS, which could manage and analyze transmitted traffic information, and it could offer handled road-flow information to Web-site in realtime.

• Journal of Navigation and Port Research
• /
• v.40 no.2
• /
• pp.57-65
• /
• 2016

The port throughput situation has changed since the 2008 financial crisis in the US. Therefore, we studied the situation, accurately estimating port traffic of Korean port after the 2008 financial crisis. We ensured the proper port facilities in response to changes in port traffic. In the results of regression analysis, Korean GDP and the real effective exchange rate of Korean Won were found to increase the container throughput in Korean and Busan port, as well as trade volume with China. Also, the real effective exchange rate of Korean Won was found to increase the port transshipment cargo volume. Based on the ARIMA models, we forecasted port throughput and port transshipment cargo volume for the next six years (72 months), from 2015 to 2020. As a result, port throughput of Korean and Busan ports was forecasted by increasing annual the average from about 3.5% to 3.9%, and transshipment cargo volume was forecasted by increasing the annual average about 4.5%.

• Journal of Korean Society of Transportation
• /
• v.13 no.1
• /
• pp.5-33
• /
• 1995

This paper addresses the outlines of the traffic signal timing principles engaged in TRACS and the results of field test. Research team, encompassing research institute, university, and electronic company, conducted the three-year project for developing the new system, named TRACS(Traffic Adaptive Control System). The project was successfully completed in 1994. TRACS aims at accomplishing the objectives of better traffic adaptability and more reliable travel time prediction. TRACS operates in real-time adjusting signal timings throughout the system in response to variations in traffic demand and system capacity. The purpose of TRACS is to control traffic on an area basis rather than on an isolated intersection basis. An other purpose of TRACS is to provide real-time road traffic information such as volume, speed, delay , travel time, and so on. The performance of the first version of TRACS was compared to the conventional TOD control through field test. The test result was promi ing in that TRACS consistantly outperformed the conventional control method. The change of signaltiming reacted timely to the variation of traffic demand. Extensive operational test of TRACS will be conducted this year, and some functions will be enhanced.

• Journal of the Society of Disaster Information
• /
• v.14 no.2
• /
• pp.230-237
• /
• 2018

Purpose: In Korea, traffic information is collected in real time as part of Intelligent Transportation System to enhance efficiency of road operation. However, traffic information based on real-time data is different from the traffic situation the driver will experience. Method: In this study, forecasts were made for future highway traffic by day and time period by adjusting the Archived data reference days to 3, 5 and 10 days based on existing traffic Archived data. Results: Fewer days of reference in the past showed smaller errors. The prediction of Monday based on five past histories showed greater errors than the 10 past histories, as the traffic flow on the sixth Monday of 2016 was somewhat different from the usual holiday. Conclution: This study shows that less of the reference days of the past history when estimating traffic volume, the more accurate the data of the traffic history of the event can be used on special days.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.4
• /
• pp.330-337
• /
• 2000

This paper proposes a new concept of coordinating green this which controls 10 traffic intersection systems. For instance, if we have a baseballs game at 8 pm today, traffic volume toward the baseball game at 8 pm today, traffic volume toward the baseball game will be incr eased 1 hour or 1 hour 30 minutes before the baseball game. at that time we can not pred ict optimal green time Even though there have smart elctrosensitive traffic light system. Therefore, in this paper to improve average vehicle speed and reduce average vehicle waiting time, we created optimal green time using fuzzy rules and neural network. Computer simulation results proved reducing average vehicle waiting time proposed coordinating green time better than electro-sensitive traffic light system. Therefore, in this paper to improvevehicle speed and reduce average vehicle waiting time, we created optiual green time fuzzy rules and neural network. Computer simulation results proved reducing average vehicle waiting time which proposed coordinating green time better than electro-sensitive traffic light system dosen't consider coordinating green time.

• Journal of Korean Society of Transportation
• /
• v.24 no.7 s.93
• /
• pp.65-79
• /
• 2006

The ATIS(Advance Traveler Information System), as one part of ITS, is a system aiming to disperse traffic volume on transportation networks by providing traffic information to transportation users on pre-trip and en-route trips. One of tools in ATIS is usage of VMS(Variable Message Signs). It provides to the drivers with direct information about state of processing direction. which is considered as the most effective method in ATIS. The purposes of providing VMS information are classified two categories. One is to provide simple information to drivers for their convenience. The other is to manage traffic demand to improve transportation network performance. However, for more effective and reliable VMS information, several strategies should be taken into account. The main VMS management strategy is "Traffic Diversion Strategy for minimum delay" when traffic congestion or incident are occurred. For effective operation. firstly. reasonable diversion traffic volume is determined by network traffic condition Secondly, it is necessary to make providing information strategy which reflects driver response behavior for controling diversion traffic volume. This paper focuses on the providing real-time route guidance information by VMS when congestion is occurred by the incidents. This sturdy estimates time-dependent system optimal diversion rate that inflects travel time and queue lengths using traffic flow simulation model on base Cellular Automata. In addition, route choice behavior models are developed using binary logit model for traffic information variable by traffic system controller. Finally, this study provides time-dependent VMS massage contents and degree of providing information in order to optimize the traffic flow.