Abstract:
Tra c congestion is a major issue in many modern cities around the world. Over many
years, di erent solutions have been proposed and tested to solve the problem. One such solution
is to build a new tra c infrastructure however this proves to be expensive and may not be
possible in some places. Another solution is to optimize the existing tra c network by smartly
controlling tra c lights at intersections, to ensure that vehicles move smoothly along transportation
routes. Analysing the situation in Pakistan, the current tra c lights in Islamabad
city, which are installed across the city, are insu cient to address the congestion issues since
they have particular pre-determined times for green and red lights. is results in large queues
leading to disruptions in the daily lives of passengers. In this research, we used xed and actuated
tra c light control methods to optimize tra c signal programs generated by SUMO,
which is an open-source microscopic simulator. By using delay time and queue length as the
parameters to reduce the congestion, one single intersection has been modelled in SUMO.
Actuated tra c light control has shown a strong potential to cater to tra c uctuations to
achieve desired objectives and e ectively reduced tra c congestion by adjusting tra c signal
plans. To test the e cacy of the used strategy, one single intersection has been used which
is controlled by an Intelligent tra c light control plan (ITCS). Further, actuated tra c light
control was found to reduce intersection delays by up to 27% relative to xed tra c light
control which can be further improved by using adaptive tra c light control strategies based
on machine learning algorithms. Moreover, Fluctuations in tra c owhave been catered be er
by actuated tra c light control than xed tra c light plan.
