Traffic speed prediction method

Abstract

The invention discloses a traffic speed prediction method, belonging to the field of intelligent transportation. The traffic speed prediction method is characterized by extracting traffic flow data bymeans of a loop detector, using a three-dimensional data matrix constructed by three attributes of flow, speed and occupancy acquired in an entire detection road section as input, performing extraction and weighting operation on spatio-temporal features by means of a convolutional neural network, and finally predicting a traffic speed in a future time period by fusing the spatio-temporal data. Through integrating the traffic data of the entire road section, the traffic speed prediction method can predict the speeds of a plurality of detection points at one time, and improves the efficiency when compared to the single point cycle prediction. In addition, through the fusion of the multi-attribute spatio-temporal features and the weighting operation, the effectiveness of traffic features iseffectively enhanced, and the precision of predicting tasks is improved.

Description

technical field [0001] The invention relates to the field of intelligent transportation, in particular to a traffic speed prediction method based on a weighted convolutional neural network model. Background technique [0002] With the development of the economy, the number of motor vehicles owned by urban residents continues to increase, and the load on the road traffic network is also continuously rising. In order to solve the traffic problems caused by high loads, intelligent transportation systems have been proposed. The most core and challenging part is the prediction of road traffic flow data, which is also the basis for realizing other advanced functions. The goal of short-term traffic flow forecasting is to predict the future traffic state index variables such as speed and occupancy rate at the location where the sensor is located. Usually, the time range is 5 to 30 minutes. Traffic flow has a very complex time-spatial dependency. In addition to the complex time seri...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem that it is difficult to effectively extract the spatio-temporal features of traffic flow in the traffic speed prediction task, and how to measure the validity of the spatio-temporal features, the present invention provides a traffic speed prediction method based on a weighted convolutional neural network model

Images