Information Technology Reference
In-Depth Information
Many studies were conducted to determine the factors that influence the choice of
the operating speed. Poe, Tarris and Mason [8] showed that access points, land-use
characteristics and traffic engineering features have influences on vehicle speed on
low speed urban streets. Haglund and Aberg [9] showed that the posted speed limit
has influence on drivers' speed choice behavior. Fitzpatrick, Carlson, Brewer and
Wooldridge [10] evaluated the influence of geometric, roadside and traffic control
devices on drivers' speed on four-lane suburban arterials and found that posted speed
limit was the most significant variable for both curve and straight sections. Wang [11]
demonstrated that the number of lanes, sidewalks, pedestrian movements, and access
density have significant influences on the drivers' behavior of choosing operating
speed. Fildes, Fletcher and Corrigan [12] and Fildes, Leening and Corrigan [13] found
that the road width and the number of lanes have the greatest influence on speed
choice. Tignor and Warren [14] showed that the number of access points and the
nearby commercial development have the greatest influences on the vehicle speeds.
Most of these studies used different model approaches range from simple linear re-
gression models to complex curvilinear regression equations [8], [11], [15]. Most of
the existing models attempt to quantify the operating speed based on physical
conditions such as road geometric design, roadside development and traffic control
devices. All of these models used 85
th
percentile speed as a representative measures
for operating speed.
No studies on the use of FLM to estimate the 85
th
percentile speed have been
found. The FLM approach has the premise to tackle the imprecise, vague and uncer-
tain relationship between the inputs and outputs. The proposed system can be re-
garded as an expert system or a knowledge base. It is critically important that the
design of such system should account for the imprecise, incomplete or not totally
reliable information [16]. The key feature of the FLM is the suitability to incorporate
intuition, heuristic and human reasoning [17] and such technique is useful for prob-
lems that entail probabilistic or stochastic uncertainty (human behavior modeling), or
problems that cannot be easily represented by mathematical modeling because of the
complexity of the process [18]. Fuzzy set theory provides a strict mathematical
framework in which vague conceptual phenomena can be precisely and rigorously
studied [19]. The word imprecise or vague does not mean the lack of knowledge of
data; rather it indicates the sense of vagueness of the value of parameters.
The objective of this paper is to develop a fuzzy logic based approach to estimate
the 85
th
percentile speed for different urban road segments based on road segments
attribute data for weekday and weekend. In doing so, four urban road segments (one
local and three arterial roads) of Al Ain city of United Arab Emirates have been se-
lected randomly (termed as 'Site 1' to 'Site 4'). Only four road segments were se-
lected because of limited time and resources for conducting the study. The authors
do recognize that the limited data collection cannot be used to make general conclu-
sions on the validity of the devised FLM for a general network. We emphasize here
that the main contribution of this study is the introduction of the concepts and
the procedure to develop the FLM that can be generalized to any network given that
adequate data collection on a representative sample size is fulfilled.
