Abstract
In this paper, a method for traffic pattern analysis and state prediction for correlated routes in the road network is proposed. First, the concepts of correlated route, correlated route chains, and correlated route sets are defined, and a route correlation degree calculation model that considers route traffic heterogeneity and its judgment criteria are proposed to determine the correlated route sets in the region. Second, we incorporate the self-organizing mapping (SOM) algorithm with Dunn index (DI), named as SOM_DI, to classify the traffic states on the correlated route chain and determine the optimal number of traffic state. The traffic pattern on the correlated path chain is analyzed to obtain the temporal state chains and the spatial state chains. Finally, an algorithm is proposed to select the input spatio-temporal features of the support vector regression (SVR) model and predict the traffic state on the correlated route chain, which is named as STFS_SVR. The simulation results show that the method proposed in this paper can accurately classify the correlated routes of regional traffic and its optimal traffic state.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call