A Traffic Data Filling Method Based on Generative Adversarial Network

Abstract

The invention relates to a traffic data filling method based on a generative confrontation network, which combines a 3D convolutional neural network with a generative confrontation network to construct a 3DConvGAN model, first uses historical data to train the 3DConvGAN model, and uses a 3D convolutional neural network The network extracts the spatiotemporal features of the data at the near moment; secondly, it is set to measure the difference between the real data of the known point and the generated data, and the optimal input of the generated network is obtained by minimizing the loss function; finally, the optimal input is used to generate The network obtains the optimal generated data to realize the recovery of traffic data. The invention overcomes the deficiency that the prior art cannot make full use of the historical information and spatio-temporal characteristics of the traffic data, fully utilizes the historical traffic data and effectively extracts the spatio-temporal characteristics of the traffic data, thereby improving the recovery accuracy of the traffic data.

Description

technical field [0001] The invention relates to the field of intelligent traffic and deep learning, in particular to a traffic data filling method based on a generative confrontation network. Background technique [0002] In the intelligent transportation system, traffic data collection is an important part. The main traffic data collection methods include: fixed-point detectors such as induction coil detectors and infrared detectors, or dynamic monitoring using devices such as global positioning systems (GPS) and mobile phone communications. However, due to problems such as equipment damage and abnormal data transmission, the collected traffic data will be missing data. The missing traffic data not only reduces the research value of the data, but also affects the subsequent research work. How to effectively restore the missing traffic data and ensure the integrity of the data plays a vital role in the work of the intelligent transportation system. [0003] So far, the ma...

AI Technical Summary

Problems solved by technology

[0004] In view of this, the purpose of the present invention is to propose a traffic data filling method based on a generative confrontation network, which can overcome the shortcomings of the prior art that cannot make full use of the spatio-temporal characteristics of traffic data and the historical information of the data, and can recover multiple traffic data by using the generative model. missing data and improve the accuracy of recovery

Images