More Smileys on the road

More Smileys on the road

van der Vorst, H.M.

Hellendoorn, J. (mentor)

Mechanical, Maritime and Materials Engineering

Delft Center for Systems and Control (DCSC)

2017-05-17

Congestion is one of the big annoyances people experience in their everyday life. Furthermore, congestion has negative effects on both the environment and the economy. In short, it forms a very big problem to society. This thesis focuses on a specific type of congestion, namely a ‘wide moving jam’. This is typically a short jam that propagates in the opposite direction of the traffic. The short jam usually originates at a ‘weak link’ of any specific road. An example of such a weak link is the location where four lanes become three lanes and people are asked to merge. The general rule is that the outflow of traffic in a wide moving jam is 30% lower than the free capacity of the road. This means that the realized capacity of the road can be much lower than the theoretical capacity. A very well-known example of this in the Netherlands is the freeway route A13-A20-A16 which passes Rotterdam. It has been shown that a large part of the congestion on the A13 is caused by moving jams [Vonk Noordegraaf, 2013]. This route will be the subject of the analysis performed in this thesis. Often there is a combination of factors that contribute to the formation of a wide moving jam. Speed and density are high but there is another specific reason (or more) at play. In such a situation the state of traffic is not stable and therefore a small disruption can cause a jam. The general strategy in combatting these wide moving jams is to lower the inflow of traffic onto the jam. Algorithms have been developed that are able to detect and dissolve wide moving jams with use of variable speed signs to control traffic flow. However, while this measure is successful in solving a wide moving jam there are side effects. The variable speed sign changes the behaviour of the road users in such a way that congestion will form elsewhere. It has been studied that the impact of the side effects are in the same order of magnitude as the initial problem. Instead of cutting off the flow to solve the moving jam why not stimulate the flow at the known locations where they originate? In this thesis it is proposed to solve the operational bottleneck that causes a wide moving jam by stimulating vehicles to maintain their speed. At the origin of a wide moving jam a dynamic road sign will be placed that displays your speed with a L if you are driving too slow and a J if your speed is sufficient. To investigate the effect of stimulating vehicles to increase or maintain their speed, traffic is divided into two groups, a fast group and a slow group. Simulations are done where vehicles from the fast group and the slow groups are mixed. The range of free speed is constant but the compositions of vehicles with high free speed and low free speed is variable. The simulations show how more uniform free speed results in less congestion in both cases, whether the ratio of fast to low vehicles is close to 100 or 0. The difference is in the level of congestion, when the majority of vehicles are fast the flow discharge rate stays approximately constant. However, when the majority of vehicles is slow, the flow discharge rate is not. As the percentage of fast vehicles is decreased congestion decreases but speed in congestion decreases as well, therefore the critical density decreases. A smiley traffic sign that encourages vehicles to increase their speed can be used in combatting wide moving jams by reducing the bottleneck and therefore the chance a wide moving jam will form. When speed and density are high a smaller disruption is less likely to cause a wide moving jam. It is difficult to define a critical size for the bottleneck as it depends on the state of traffic. However, it is found that when the percentage of fast vehicles succeeds the percentage of slow vehicles the speed inside the congestion is more stable.

http://resolver.tudelft.nl/uuid:1d27cf52-7052-4550-9332-fb6aca85206f

Student theses

master thesis

(c) 2017 van der Vorst, J.