Predictive Modelling of Traffic Delay in Arterial Road Based on Machine Learning Algorithm

Traffic delay in arterial road is an important criterion in the judgment of service level of transportation. At present, there exist a certain degree of congestion problems in most road intersections, especially in the holidays. In order to assess and predict the degree of congestion in road intersection, the traffic simulation research of a chosen area in Jinshan road was undertaken. PTV VISSIM was used to generate the simulated traffic flow data. Linear Regression model, SoftMax and machine learning were combined to process the simulated data generated from VISSIM micro simulation. This is in the aim of building a proper model that can be used for data generation like predicted transport traffic delay as well as predicted degree of congestion under ranged traffic volumes. This research may act as a guidance and reference for city designers and transportation engineers in their further research about Jinshan Road as well as others.