In South Korea, the increasing number of cars has caused various traffic issues such as frequent road congestion and air pollution. To solve these problems, local governments have attempted to manage traffic based on intelligent transport system (ITS) and advanced traffic management systems (ATMS). However, the system was built simply by expanding the infrastructure and, eventually, it decreased efficiency of traffic management. For fundamental solution to the problems, it is necessary to find a way to maximize efficiency of existing traffic management. Among various traffic issues, signal operation is one of the most important factors that influence road congestion and the conventional signal operation did not take into account various traffic patterns and, therefore, resulted in low efficiency as, for instance, a small traffic flow receives a long green and causes another traffic delays. Flow in cities is related to various traffic patterns as particular groups of fleets are formed in different times, and it seems necessary to design a signal operation system for different situations.
This study was conducted to design a signal control system to improve efficiency in road operation on the network level by classifying situational traffic patterns based on a short-term traffic prognosis data and enabling fluid signal operation.
For situational signal control based on traffic prognosis data, traffic patterns were established in subject areas based on traffic demand forecasting models of local governments and these patterns were classified into five traffic situations (inactive flow, normal flow, crowded flow-entering the city, crowded flow-in the city, and incident situation flow). Also, main arterial were selected and the relevant directional design hourly volume (DDHV) that need to be considered as priority in signal operation was reviewed based on the upward and downward traffic volumes and V/C. And then, a critical intersection (CI) were selected as they are the basis of optimization and interlocking of signalized intersections within the selected arterial. Local optimization was performed on the CI to calculate the optimal cycle and phase and set the main sub-area (SA). By performing network optimization of situational networks for the interlocked group, frame signals were created for the aforementioned five traffic situations (signal program: S0, S1, S2, S3, S4, S5). The frame signals determine detector data and existing maintenance time conditions based on signal convert program, and, if all relevant conditions are met, change to appropriate signals according to the priority. Here, the signal convert program was modularized by using a COM interface named Visual Basic Application.
In this study, which classified traffic situations based on short-term traffic prognosis data to build a system to control signals, frame signals were set by optimizing network for five different traffic situations, and the frame signal of each situation enabled understanding effect of DDHV delay improvement and also, based on the detector data, changing signals according to the situation.