Intelligent and Predictive Vehicular Networks

• 作者: Constantinos Ierotheou Maryam Roshanaei Richard Anthony Schmidt Shilukobo Chintu
• 来源期刊: 智能控制与自动化(英文)
• 学科: 医学
• 关键词: Simulation of Urban Mobility SUMO Duarouter Fully Balanced Multiple-Candidate Shortest Paths (FBMKP) E1 Induction Loop E3 Detector Re-Routing Braess PARADOX TRAFFIC CONTROL INTELLIGENT (TraCI) Partially Re-Routed Shortest Path Method TRAFFIC Light CONTROL FBMKP
• 年，卷（期）: 2014,（2）
• 页码范围: 60-71
• 页数: 12页
• 分类号: R73

摘要

Seeking shortest travel times through smart algorithms may not only optimize the travel times but also reduce carbon emissions, such as CO2, CO and Hydro-Carbons. It can also result in reduced driver frustrations and can increase passenger expectations of consistent travel times, which in turn points to benefits in overall planning of day schedules. Fuel consumption savings are another benefit from the same. However, attempts to elect the shortest path as an assumption of quick travel times, often work counter to the very objective intended and come with the risk of creating a “Braess Paradox” which is about congestion resulting when several drivers attempt to elect the same shortest route. The situation that arises has been referred to as the price of anarchy! We propose algorithms that find multiple shortest paths between an origin and a destination. It must be appreciated that these will not yield the exact number of Kilometers travelled, but favourable weights in terms of travel times so that a reasonable allowable time difference between the multiple shortest paths is attained when the same Origin and Destinations are considered and favourable responsive routes are determined as variables of traffic levels and time of day. These routes are selected on the paradigm of route balancing, re-routing algorithms and traffic light intelligence all coming together to result in optimized consistent travel times whose benefits are evenly spread to all motorist, unlike the Entropy balanced k shortest paths (EBkSP) method which favours some motorists on the basis of urgency. This paper proposes a Fully Balanced Multiple-Candidate shortest path (FBMkP) by which we model in SUMO to overcome the computational overhead of assigning priority differently to each travelling vehicle using intelligence at intersections and other points on the vehicular network. The FBMkP opens up traffic by fully balancing the whole network so as to benefit every motorist. Whereas the EBkSP reserves some routes for cars on high priority, our