eunite

research

applications

knowledge transfer

e-archives



Installation and experiences of field testing a fuzzy signal controller

Submitted by Josefa Z. Hernandez 27.11.2001, IBA B

Problem

Signal controllers at real intersections of different road or streets have different things to realise: (i) the high volumes of traffic during peak hours, (ii) pedestrianís crossings, (iii) bus traffic, etc. Some of the items of a good controller are: (1) Vehicles donít have to wait a lot of time at the intersection in order to there wonít be long queues, (2) Pedestrianís waiting times have to be shorter and (3) Busís travel times donít have to be increased.

Solution

At the Helsinki University of Technology, a fuzzy signal controller has been used at a real intersection with two different algorithms: a normal fuzzy (Fu) and a multi-objective fuzzy (Fm). The Fm algorithm was designed for safety and environmental aspects as well as efficient flow, whereas the Fu algorithm only considered the traffic fluency. The fuzzy-control algorithm works at two levels. The upper level classifies the traffic situation in oversaturated, normal or low-demands conditions. The lower level adjusts the green and cycles time.

Status and results

The fuzzy signal controller was installed in a real intersection. Results were based on simulations and field measurements, and both indicated that the fuzzy controller was very competitive against traditional vehicle-actuated control, if traffic volumes were higher than low-demand. Some of the advantages of the fuzzy-signal controller were: the average delays were approximately 3-8 seconds shorter, the percentages of stops were 2-12% lower, the bus delays were smaller in 8/9 cases, and there were good savings in fuel and emissions. However, the better traffic fluency was only one advantage. Pedestrians can also benefit; because the cycle time was on average 8 seconds shorter.

Adaptivity and portability

One of the benefits of the fuzzy logic lies in its ability to handle linguistic information by representing its as a fuzzy set. Changing these fuzzy sets and the fuzzy rules this controller can use for different intersection which different traffic conditions.

More information

J. Niittymaky. European Journal of Operational Research 131 (2001) 273-281

technical reports & case studies

executive summaries

scientific papers

tutorials

bibliography

e-dictionary

search eunite with

google



  home