Uvodnik
https://doi.org/10.7307/ptt.v35i2.34
Study of Full Controlled Green Time Roundabouts – An Intelligent Approach
Hong Ki An
; Faculty of Civil Engineering and Technology, Universiti Malaysia Perlis
Gimik Bae
; School of Civil & Environmental Engineering, Daejin University
Dong Sun Kim
; School of Civil & Environmental Engineering, Daejin University
Sažetak
When roundabouts face congestion problems, the transition to signalised roundabouts is considered a solution to the problem. The majority of studies have concentrated on how to calculate the optimal cycle length and signal timing to minimise congestion at roundabouts. To date, intelligence algorithms with multi-objectives such as queue length, number of stops, delay time, capacity and so on are widely used for calculating signal timing. Although roundabout congestion can be generated by the weaving zone reducing roundabout capacity, there have been minimal studies which take into account the density in the weaving zone. This study proposed a hybrid gravitational search algorithm – ABFO random forest regression with the following objectives: density, delay time and capacity to find the optimal cycle length and green time in each phase of Changwon city hall roundabout in South Korea as a case study. The optimal cycle length and green time were calculated in MATLAB and microscopic simulation VISSIM sought the effectiveness of a signalised roundabout. The result of the analysis demonstrated that signalised roundabouts with 102 seconds cycle length (phase 1 – 65 seconds of green time and phase 2 – 37 seconds of green time) can reduce density by 46.1%, delays by 32.8% and increase roundabout capacity by 14.8%.
Ključne riječi
signalised roundabout; green time; optimisation; VISSIM; hybrid GSA-ABFO algorithm
Hrčak ID:
301141
URI
Datum izdavanja:
25.4.2023.
Posjeta: 244 *