TY - GEN
T1 - Analysis of the influence of traffic signalization using a stochastic algorithm in the reduction of queues and delays at intersections with high traffic flow
AU - Urbano, Jhonan
AU - Bassini, Fernanda
AU - Silvera, Manuel
AU - Campos, Fernando
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - The constant increase of vehicular demand at intersections hurts intersection crossing times and queue formation on avenues. This paper presents a microsimulation model using a stochastic algorithm for traffic signal control based on two variables: queue length formation and crossing time delays. To address this problem, a stochastic algorithm is built using Python software with the total lengths of each traffic light cycle and the two variables to be solved (queue formation and crossing delays) as parameters, and the number of iterations to be performed will be included in the algorithm. These variables will be used as key indicators to obtain the green and red-light duration times for each traffic light at both intersections. To validate the effectiveness of the proposed model, different traffic simulations are performed in the intersection section using Vissim 9.0. Using this microsimulation software, it was possible to recreate the behavior of the vehicles that were analyzed using a filmographic record. Different iterations were used to determine the trend of improvement in the model using the results of the algorithm through different phase diagrams used in the microsimulation. By using the first iterations of the times generated by the algorithm, a considerable improvement in the performance of the traffic light control system was observed. A notable decrease in crossing times has been achieved, with reductions ranging between 5% and 9%. In addition, a considerable decrease in queuing has been observed, with a reduction ranging from 20% to 34%.
AB - The constant increase of vehicular demand at intersections hurts intersection crossing times and queue formation on avenues. This paper presents a microsimulation model using a stochastic algorithm for traffic signal control based on two variables: queue length formation and crossing time delays. To address this problem, a stochastic algorithm is built using Python software with the total lengths of each traffic light cycle and the two variables to be solved (queue formation and crossing delays) as parameters, and the number of iterations to be performed will be included in the algorithm. These variables will be used as key indicators to obtain the green and red-light duration times for each traffic light at both intersections. To validate the effectiveness of the proposed model, different traffic simulations are performed in the intersection section using Vissim 9.0. Using this microsimulation software, it was possible to recreate the behavior of the vehicles that were analyzed using a filmographic record. Different iterations were used to determine the trend of improvement in the model using the results of the algorithm through different phase diagrams used in the microsimulation. By using the first iterations of the times generated by the algorithm, a considerable improvement in the performance of the traffic light control system was observed. A notable decrease in crossing times has been achieved, with reductions ranging between 5% and 9%. In addition, a considerable decrease in queuing has been observed, with a reduction ranging from 20% to 34%.
KW - algorithm
KW - delays
KW - queuing
KW - semaphores insert
KW - stochastic
UR - http://www.scopus.com/inward/record.url?scp=85179556946&partnerID=8YFLogxK
U2 - 10.1109/CONIITI61170.2023.10324268
DO - 10.1109/CONIITI61170.2023.10324268
M3 - Conference contribution
AN - SCOPUS:85179556946
T3 - 2023 9th International Conference on Innovation and Trends in Engineering, CONIITI 2023 - Proceedings
BT - 2023 9th International Conference on Innovation and Trends in Engineering, CONIITI 2023 - Proceedings
A2 - Triana, Jenny Paola Hernandez
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 9th International Conference on Innovation and Trends in Engineering, CONIITI 2023
Y2 - 4 October 2023 through 6 October 2023
ER -