Editorial

Special Issue: Network Reliability and Transport Modelling

William H.K. Lam, Guest Editor

With increasing demands for better and reliable service due to improvements in the quality of life, many systems (such as the communication networks and drainage systems) have incorporated "reliability analysis" as an integral part in their planning, design and operation. However, very little attention has been given to the reliability analysis of road networks. On the other hand, Intelligent Transport System (ITS) has developed rapidly in recent years. Impetus was received from the investigation of the effects of ITS on two types of road congestion; namely, recurrent and non-recurrent congestion. The former is referred to as the predictable traffic conditions caused by an excess of demand over supply, following a defined pattern, e.g. peak period traffic and seasonal variations. The latter includes non-recurrent congestion situations resulting from a temporary reduction in the capacity of a length of road due to accidents, bad weather, construction works or other short-term disruptions to traffic flow. The introduction of ITS and the concept of "reliability analysis" has generated an urgent need for better understanding of road network reliability under recurrent and non-recurrent congestion scenarios.

In the International Workshop on Network Reliability and Transport Modelling, held in Hong Kong in December 1998, experts from Australia, Canada, Hong Kong, Japan and UK have presented papers on network reliability and transport modelling. Among them, a number of papers are focused on the subject of road network reliability which is of particular interest to practicing engineers and researchers involved in transport network design and analysis. These papers are worthy of further elaboration and expansion. This Special Issue is devoted to the subject of network reliability and to provide a broader platform for dissemination of the research findings by these authors. In total, eight papers are included in this issue and they are presented as follows.

In the first paper, Iida provides a comprehensive review of the basic concepts of road network reliability. Future directions of network reliability are also discussed on various aspects for practical use. In the second paper, Bell presents a game theoretical approach for assessing the robustness of transport networks. A hypothetical game is envisaged between utility maximising network users on the one hand and a utility minimising "evil entity" on the other. It is shown that the mixed strategy Nash equilibrium for this game offers a useful measure of network reliability, as it is the utility that would prevail if network users were extremely pessimistic.

The third paper of this issue is on evaluation of network reliability using a logit-based stochastic user equilibrium (SUE) approach. Asakura proposes an evaluation process for measuring the performance reliability of a road network with some closed or partially degraded links. A multi-class SUE model is used to assess the effects of providing travel information in a network with two groups of road users: informed and non-informed drivers. It is found that providing information generally increases network performance reliability in terms of the probability of connectivity of an origin-destination (O-D) pair or the probability that the travel time between an O-D pair is within an acceptable limit.

In the fourth paper, Lam and Xu use a probit-based SUE approach to develop a traffic flow simulator (TFS) for assessing the road network reliability in terms of travel time reliability. A more attractive approach of the TFS is to estimate link flows and to update the O-D matrix in one stage on the basis of prior O-D demand and partial traffic count data. Moreover, it can also estimate link and path travel times together with their variance and co-variance.

In the fifth paper, Chen et al introduce the capacity related reliability for transportation networks with random link capacity due to disturbances such as accident and bad weather. A Monte Carlo simulation procedure is used to estimate the capacity related reliability which is defined as the probability that the road network can accommodate a certain level of traffic demand, and is built on the concept of network reserve capacity.

In the sixth paper, Wong and Yang present an iterative scheme for a combined signal optimisation and assignment problem in a road network. The signal settings are optimised by means of a group-based technique while a path-based assignment algorithm is used to solve the traffic assignment problem. For the evaluation of network performance by optimising the signal settings, a performance index of the road network is used and considered as the total travel time in vehicle-hours/hour. In order to examine the convergence of the proposed approach and to test the reliability of the performance index, a random offset method is employed in which the iterative scheme is applied to a number of starting points each with a distinct set of initial signal offsets at intersections. It is found that the iterative scheme is always able to produce good results even with different starting points. Moreover, the probability distribution of the optimised performance index can also be obtained for the whole study network.

Taylor contributes the last two papers to dense network modelling with two main themes. The first one discusses the changing nature of traffic management technology and the underlying objectives behind traffic management practice, while the second one is concerned with the use of measures of network reliability in models particularly as an element of the evaluation of alternative network configurations.

A better understanding of the network reliability will enable us to design more economical and efficient road networks and to ensure the reliability of these road networks under re-current and non-recurrent congestion scenarios. In conclusion, the editor hopes this issue will bring the subject of network reliability to the attention of the researchers. I hope it that it will lead to the advancement of modelling techniques and solution algorithms, as well as lead to practical applications of network reliability concepts.

Back to Vol. 33, No 2 Listing
Return to Journal Listing	Return to HomePage