This paper presents data, collected from video-recording, on the microscopic details of merging and weaving manoeuvres under congested traffic conditions. Based on these observations, a classification of the manoeuvres into free, forced and cooperative lane changes is proposed. A new lane change model is developed, incorporating explicit modelling of vehicle interactions using intelligent agent concepts. The model was implemented in the ARTEMiS traffic simulator, and several hypothetical test studies were conducted to demonstrate the capabilities of the new model. The results show that the model is able to reproduce the observed behaviour of individual vehicles in terms of speed, gap acceptance and conflict-resolution in all three types of lane change manoeuvres, and hence, it is able to simulate highly congested flow conditions in a realistic manner. The macroscopic results in terms of speed-flow relationship are close to the typical expected results. The model can simulate both freeways and signalised urban arterial networks. (C) 2005 Elsevier Ltd. All rights reserved.
