Road condition prediction method and road condition prediction system based on space-time relationship

The invention discloses a road condition prediction system and a road condition prediction method based on a space-time relationship. The road condition prediction system based on the space-time relationship comprises: a section space influence degree determination part, which determines the space influence degree of each section to be predicted among a plurality of sections influenced by surrounding sections; a road condition prediction model establishing part, which establishes a road condition prediction model for the section to be predicted by utilizing the determined space influence degree and the historic road condition data of the plurality of sections; and a road condition prediction part, which predicts a road condition of the section to be predicted within a future time period by utilizing real-time road condition data and the road condition prediction model. The invention further discloses a road condition space influence degree determination device and a road condition space influence degree determination method, and a road condition prediction device and a road condition prediction method. In the invention, by taking the space influence degree of the section as a space operator and combining a time series model, an influence on the current section from the surrounding sections on a plurality of space orders can be taken into consideration, and the road condition of a space range is measured more realistically, thereby being favourable for improving prediction accuracy.