Multidirectional traffic flow prediction method based on interest point space-time residual neural network

Abstract

The invention discloses a multidirectional traffic flow prediction direction based on a point-of-interest space-time residual neural network. According to the method, based on short-time vehicle trajectory data, time signals and interest point signals are added to enhance spatial-temporal features, 3D CNN is utilized to extract spatial-temporal features of traffic flow changes along with time, a residual neural network is combined to avoid model overfitting, finally, weighted compression is performed on information with the spatial-temporal features, and a traffic flow distribution matrix with a moving state is output. And the prediction of the regional traffic flow is realized. According to the method, the characteristic that the space distribution of the traffic flow changes along with time is well associated, the relation between the time signals and the interest point signals is integrated, the method has the advantages of being high in precision, high in practicability and the like, and decision support can be provided for public facility deployment, traffic diversion, land use planning and the like.

Description

technical field [0001] The invention relates to the field of traffic flow forecasting, in particular to a multi-directional traffic flow forecasting method based on interest point space-time residual neural network. Background technique [0002] Traffic data, as a ubiquitous spatiotemporal data type, exhibits correlations in both temporal and spatial dimensions. In addition, traffic forecasting research in the past few decades has mainly focused on taxis, private cars, and shared bicycles, while less research has been done on electric bicycles. In addition, due to the short driving time and short driving distance of electric bicycles, the periodicity of electric bicycle trajectory data is not obvious. Various reasons make the prediction of electric bicycles more challenging than traditional traffic prediction. In recent years, electric bicycles have gradually become an indispensable means of transportation in people's daily life due to their low price and small size. Accor...

AI Technical Summary

Problems solved by technology

[0005] 2. Temporal correlation: Observations obtained at consecutive time intervals are highly correlated, i.e., if an upward trend is detected from observations in the last few consecutive time intervals, there is a high probability that the observations at the next moment will increase

Images