Probability modelling of pedestrian collision at signalised intersections Available to Purchase

The behaviour of pedestrians and drivers at signalised intersections are analysed. Three pedestrian safety evaluation methods including “accident analysis technique”, “traffic interference technique” and “traffic condition evaluation” are reviewed and compared. Simulation of movement in real conditions at intersections is done and the results are compared. Results show that time of day is an important factor that affects the behaviour of drivers. Evening driving reduces the minimum distance and minimum ‘time to crash’ (TTC), implying morning driving is less dangerous than driving in the evening. Vehicle movement and pedestrian-pedestrian movement only affect the minimum distance and minimum TTC. In addition, the pedestrian-passerby visibility factor is investigated to investigate the effects on drivers' behaviour. Based on the results, the average RSE is 1.23 m, which means that the average difference between the minimum distance predicted and the minimum distance obtained is 1.23 m. The R square for the model is 0.922, which indicates that 92.2% of the variation in the experimental data can be explained by the variation in the prediction results. Based on this, the results showed that the prediction model of pedestrian and vehicle interference has a good predictive performance.