Research On The Mobility Reliability Of Urban Road Network

Urban road network, as the carrier in the transportation, is an important infrastructure of urban transport. In recent years, the traffic situation is getting worse,which makes network reliability study attracting more and more people’s attention. As a dynamic probability index, urban road network reliability is an important indicator of the conditions of the road network. The study of the urban road network reliability will provide relevant parameters for the transport planning researcher, helping decision-makers to develop and propose better solutions of transportation planning.On the basis of reading related research and articles of road network reliability issues at home and abroad, the definition of road network reliability and the analysis of the differences and relations between road network reliability and road network vulnerability, the road network reliability index will be divided into three parts:survival indicators, stability indicators and effectiveness indicators. This paper focused on unblocked reliability index. Through the analysis and characterization of the randomness of traffic supply and demand, and the analysis of the evaluation criteria of road network unblocked reliability, learning from general network reliability theory,the unblocked reliability computational model for road, path, OD pairs and entire network system is proposed. Taking the relevant sections of the units into account,using PENT algorithm to improve the computational model, and take an example to verify it. Through the study of the links between reliability indexes, it is analyzed that the research of validity indicators should be premised on stability indexes, when unblocked reliability is measured by speed, unblocked reliability and travel time travel time reliability can be converted. At last, applying the unblocked reliability to the road network design problems, establish the bi-level programming model based on unblocked reliability, use this model to solve a simple example, test the effectiveness of the model and algorithm, and lay a foundation for deeper research.