A comparative study of stochastic `least-time' path algorithms in the context of the Nottingham urban network

This paper presents experimental results and analysis of the `actual' performance of two currently implemented stochastic 'least-time' path algorithms in the context of the Nottingham urban network. The suitability of each algorithm is judged in the context of the Nottingham urban network. The particularities and availability of the input data play a pivotal role in the assessment and evaluation of the two algorithms; as the quality of the optimal path solution is clearly dependent on the accuracy of the links' traversal times. However, the data sources being used in order to estimate and predict the links' traversal times are limited. For the time being, SCOOT data (information regarding the presence of vehicles, primarily used for traffic management) is the only source of available traffic information, in the context of the Nottingham urban network. Therefore, the characteristics of the traversal limes of links, which are equipped with SCOOT inductive loop detectors, are explored. Following experimental res