Traffic data restoration method based on graph convolution time sequence generative adversarial network

Abstract

The invention discloses a traffic data restoration method based on a graph convolution time sequence generative adversarial network. The method comprises the following steps: obtaining an original traffic data set collected by traffic equipment, and carrying out abnormal value processing on the obtained original traffic data set by employing a unary Gaussian distribution outlier screening method;selecting a data set in a period of time from the data set subjected to abnormal value processing as a complete real data set, and randomly deleting the real data set according to different proportions to obtain a plurality of to-be-restored data sets; constructing a generative adversarial network model with restored traffic data by utilizing a generative network and a discrimination network; inputting the to-be-restored data sets into the generative network to obtain a reconstructed data set; then inputting the reconstructed data set and the real data set into the discrimination network together to complete dynamic adversarial training of the generative network and the discrimination network, so that the discrimination network cannot distinguish the reconstructed data set and the real data set; and carrying out traffic data restoration on the trained generative adversarial network.

Description

technical field [0001] The present invention relates to the technical field of intelligent transportation, and more specifically, relates to a traffic data restoration method based on graph convolution sequence generation confrontation network. Background technique [0002] With the development and application of urban intelligent transportation systems, urban traffic data such as radio frequency identification (Radio Frequency Identification, RFID) automatic number plate recognition data (Automatic Number Plate Recognition, ANPR), global positioning system (Global Positioning System, GPS) data, coil data, A large amount of data such as mobile phone signaling has been collected, which makes up for the shortcomings of traditional resident traffic surveys that consume a lot of manpower and financial resources, time and cost, and have low timeliness and accuracy. These data sets capture the basic state and dynamic information of the transportation network and the entire system,...

AI Technical Summary

Problems solved by technology

This makes the previous method simply repair the missing traffic data with a relatively simple adjacent intersection correlation mode in the process of data restoration, and cannot capture and represent the relevant basic traffic parameters closely related to intersections in the road network map, resulting in There is no good repair ability in the case of sudden changes in traffic conditions

Especially for scenarios with a high data missing rate, the reconstruction of the general time series interpolation method is basically ineffective

Images