DOI QR코드

DOI QR Code

A Study on the Introduction of Bus Priority Signal using Deep Learning in BRT Section

BRT 구간 딥 러닝을 활용한 버스우선 신호도입 방안에 관한 연구

  • 임창식 (도로교통공단 부산지부) ;
  • 최양원 (영산대학교 드론교통공학과)
  • Received : 2019.10.28
  • Accepted : 2019.11.12
  • Published : 2020.02.01

Abstract

In this study, a suitable algorithm for each BRT stop type is presented through the network construction and algorithm design effect analysis through the LISA, a traffic signal program, for the BRT stop type in the BRT Design Guidelines, Ministry of Land, Transport and Maritime Affairs, 2010.6. It was. The phase insert technique is the most effective method for the stop before passing the intersection, the early green technique for the stop after the intersection, and the extend green technique for the mid-block type stop. The extension green technique is used only because it consists of BRT vehicles, general vehicles and pedestrians. Analyzed. After passing through the intersection, the stop was analyzed as 56.4 seconds for the total crossing time and 29.8 seconds for the delay time. In the mid-block type stop, the total travel time of the intersection was 40.5 seconds, the delay time was 9.6 seconds, the average travel time of up and down BRT was 70.2 seconds, the delay time was 14.0 seconds, and the number of passages was 29.

본 연구는 딥러닝 기술을 적용한 스마트교차로의 부산 해운대로 BRT 구간 버스정류장 유형을 대상으로 교통신호 프로그램인 LISA를 통해 네트워크 구축 및 알고리즘 설계 효과분석을 통해 버스정류장 유형별로 적합한 알고리즘을 제시하였다. 교차로 통과 전 정류장은 Phase insert 기법, 교차로 통과 후 정류장은 Early green 기법, 미드블럭형 정류장은 Extend green 기법이 가장 효과적인 것으로 분석되었고, 버스 및 일반차량과 보행자 현시로 구성하였기 때문에 Extend green 기법으로만 분석하였다. 교차로 통과 전 정류장은 교차로의 전체 통행시간은 57.8초, 지체시간은 33.2초, BRT 상·하행 평균 통행시간 85.3초, 지체시간 31.1초, 통과대수는 28대로 분석되었고, 교차로 통과 후 정류장은 교차로의 전체 통행시간은 58.2초, 지체시간은 31.8초 BRT 상·하행 평균 통행시간 102.2초, 지체시간 42.5초, 통과대수 26대로 분석되었다. 미드블럭형 정류장은 교차로의 전체 통행시간은 42.5초, 지체시간은 11.2초, BRT 상·하행 평균 통행시간 74.2초, 지체시간 17.0초, 통과대수 28대로 분석되었다. 분석결과를 토대로 버스우선 신호시범도입, 보행자 시거확보를 위한 계단식정지선, 속도감속을 위한 고원식횡단보도, 딥러닝 기술을 활용한 무단횡단금지 경고 벨 및 VMS 설치 등으로 BRT 구간에서의 교통사고 감소 효과가 기대되며, 이를 확대 도입할 필요가 있다.

Keywords

References

  1. Busan Metropolitan City (2018). Final report on service quality evaluation for smart intersection traffic management system in Busan metropolitan city (in Korean).
  2. Busan Metropolitan City (2019). Final report on service quality evaluation for smart intersection traffic management system in Busan metropolitan city (in Korean).
  3. Cox, D. R. (1958). "The regression analysis of binary sequences whit discussion." Journal of the Royal Statistical Society. Series B (Methodological), Vol. 20, No. 2, pp. 215-242. https://doi.org/10.1111/j.2517-6161.1958.tb00292.x
  4. Han, M. J. and Lee, Y. I. (2006). "Establishment of bus priority signal in real-time traffic signal control." Korean Society of Transportation, Vol. 24, No. 7, pp. 101-114 (in Korean).
  5. Hong, K. S., Jung, J. H., An, K. H. and Lee, Y. I. (2011). "A study on the active transit signal control algorithm based on bus demand using UTIS." Korean Society of Transportation, Vol. 29, No. 6, pp. 107-116 (in Korean).
  6. Jung, Y. T. (2018). "A study on traffic management system by smart intersection using deep learning technology." Korea Local Information Research & Development Institute, Vol. 113, pp. 15-18 (in Korean).
  7. Kang, Y. S. (2014). A study on the BRT priority signal operation considering bus stop characteristics. Master's Degree, Kongju National University (in Korean).
  8. Lee, B. G., Lee, C. K., Yoon, I. S. and Kim, Y. S. (2012). "Development of vehicle queue length estimation model using deep learning." The Journal of the Korea Institute of Intelligent Transportation System, Vol. 11, No. 2, pp. 21-28 (in Korean). https://doi.org/10.12815/kits.2012.11.2.021
  9. Lee, Y. J., Hwang, J. S., Kim, S. H. and Lee, C. K. (2018a). "Development of vehicle queue length estimation model using deep learning." The Journal of the Korea Institute of Intelligent Transportation System, Vol. 17, No. 2, pp. 39-57 (in Korean).
  10. Lee, Y. J., Sim, M. G., Lee, S. S. and Lee, C. K. (2018b). "Development of vehicle queue length estimation model using deep learning." The Journal of the Korea Institute of Intelligent Transportation System, Vol. 17, No. 4, pp. 54-62 (in Korean).
  11. Minister of Land, Transport and Maritime Affairs (2013). Korea Highway Capacity Manual (in Korean).
  12. The Minister of Land, Transport and Maritime Affairs (2013). Korean highway capacity manual (in Korean).