Urban Road Travel Time Prediction based on ELM

The Travel Time Prediction (TTP) is an important element in the study of the advanced transportation guidance system and control system. In this paper, an advanced method with Extreme Learning Machine algorithm(ELM) has been discussed by analyzing the various travel time prediction method. The feasibility and advantages of Extreme Learning Machine in travel time prediction has been studied, and the comparisons between ELM and SVR and BPNN has been detaily discussed. Moreover, the traffic data, which is collected from REGIOLAB-DELFT platform, have been used for validation. The results show that the ELM algorithm outperforms the related value to those by SVR and BPNN.