http://www.paper.edu.cn
-1-
Analysis on Penetration of ATIS in Stochastic Dynamic
Traffic Network
1
Ren Hualing, Gao Ziyou, Long Jiancheng
State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, P. R.
China (100044)
E-mail:renhualing@jtys.bjtu.edu.cn
Abstract
The penetration of the travelers with ATIS services is an important factor that controls the benefits of
ATIS. Studies show that, the congestion extent of the urban traffic system will not always decrease
through the increasing of the market penetration of ATIS. In this paper, a bi-level programming model is
presented to determine the optimal service price and the market penetration of ATIS, in which the lower
problem is a stochastic dynamic user optimal model. A chaotic approach is used to solve the bi-level
programming model and the numerical results show the application of the bi-level programming model.
Numerical experiments show the effects of the penetration of ATIS on the the travellers’ behaviors and
the congestion level of the traffic network.
Keywords: market penetration of ATIS, bi-level programming, stochastic dynamic user optimal, chaotic
approach
1 Introduction
In the research of ATIS, one important factor is the market penetration of ATIS in urban dynamic traffic
system, that is, the rate of equipped travellers in all the travellers. Emmerink et al (1995) pointed out that
the benefit from ATIS is nearly relative to the market penetration of ATIS. Yang (1998) proposed a
mixed equilibrium model to describe the relation between the market penetration of ATIS and the benefit
of travellers.
In general, the travellers buy and equip the AIIS to get more exact traffic information and saving their
travel cost. That is to say, whether the travellers buy ATIS or not lies on whether this system is of benefit
to them, where the benefit is the travellers’ saving of travel cost during their trip. If the saving of travel
cost exceeds the purchasing cost of ATIS, then more and more travellers will buy this system, that is, the
penetration of ATIS will increase. However, Hall (1996) pointed out that the congestion extent of the
whole network is not continually decreasing along with the increasing of the penetration of ATIS; it is
relieved only in some range. If too many travellers buy this system, the congestion still cannot be relieved.
To the travellers, if the saving of travel cost is less than the purchasing cost of ATIS, then less of them
will buy this system and the penetration is decreasing. We can say that the penetration of ATIS is a
variable, and it is relevant to the saving of travel time. In a word, the determination of the market
penetration of ATIS is a researchable problem.
Based on the above analysis, we should consider this problem from two aspects: one is that the traffic
management department will consider not only the whole network when deciding the market penetration
of ATIS, but also the travellers’ reaction to its decision; the other is that the travellers choose their
dynamic travel routes according to their will or benefit under the decision of the traffic management
department, and, at the same time, their travel behaviours affect the decision of the traffic management
department. So we present a bi-level programming to optimize the market penetration of ATIS. In this
programming, the management department makes its decision to relive the network congestion, while the
travellers adjust their travel behaviours to satisfy some dynamic user optimal conditions.
1
Support by grants from the National Basic Research Program of China (2006CB705500), the National
Natural Science Foundation of China (70631001, 70801004), and the Specialized Research Fund for the
Doctoral Program of Higher Education (20050004005).