Study of pedestrian-vehicle interaction behaviour by microscopic simulation modelling

Walking is healthy, environmentally beneficial and sustainable to human society. Travellers increasingly are being encouraged to walk more. However, pedestrians’ interaction with motorised vehicles is a major constraint to their movement. Many innovative treatments have been developed to balance the two modes. Proper methods are required to evaluate and compare performances of different treatments to support decision making. Micro-simulation is a useful supplementary tool for such evaluation and comparison studies for its cost-effectiveness and non-intrusiveness. However, there is a significant gap between capabilities of existing simulation models and practical needs. New understandings of the Pedestrian-Vehicle Interaction (PVI) behaviour and corresponding micro-simulation models are required to conduct micro-simulation studies of the interaction process between the two modes to derive new knowledge of the mixed traffic.This dissertation presents the development and application of a micro-simulation model, PVISIM (Pedestrian-Vehicle Interaction SIMulation), to study PVI behaviour in a range of circumstances in an urban street environment. Key contributions relate to the collection of a substantial data base, development and validation of the model, an appreciation of the value of the approach and new understandings of PVI behaviour. A series of studies to measure behaviour based on the data collected in Beijing, China have been detailed. Intra vehicle and pedestrian behaviour models were developed and validated separately, incorporating the best available understandings from existing published studies and in accordance with the specific local data. The two modes were integrated by interpreting new findings from the study of microscopic interaction behaviour of the two modes. The complete model was validated against field data independent of those used in model development, covering a number of typical scenarios, including both unsignalised and signalised situations.The validated model was applied to study a typical unsignalised scenario by analysing system performances under different combinations of vehicular traffic and pedestrian crossing demand, in terms of efficiency, safety and environmental impact. Also, operations of different treatments including no-control, Zebra crossing, fixed-time signal crossing and Puffin crossing at two typical types of locations were compared. Interpretations and recommendations were given for each application. The results can be used to supplement existing guidelines for pedestrian related problems, and also contribute to the knowledge base to incorporate pedestrians into current micro-simulation tools in a more realistic way