Traffic big data analysis and prediction system and method based on multi-task learning

Abstract

The invention discloses a traffic big data analysis and prediction system and method based on multi-task learning. The system comprises a data acquisition terminal, a big data analysis and prediction platform and a user terminal. The data acquisition terminal comprises a data acquisition module and a data transmission module; the big data analysis and prediction platform comprises a preprocessing module, a traffic big data space-time modeling module and a multi-task learning prediction module; and the user terminal comprises a data receiving module and a user matching module. The collected vehicle-mounted GPS data and trajectory data are analyzed and modeled, and the trend of the traffic big data in a future time period is predicted, so that travel suggestions are provided for users, and the travel waiting time of the users is shortened. According to the method, the prediction accuracy of the traffic big data is improved, and the travel speed and departure time of the user can be predicted in advance, so that the traveler can reasonably arrange the departure time and avoid peak travel or go to the destination to the greatest extent, and the user can conveniently formulate a coping scheme in advance.

Description

technical field [0001] The present invention mainly relates to the field of traffic big data, in particular to a traffic big data analysis and prediction system based on multi-task learning. Background technique [0002] With the increasing number of motor vehicles in the city, the traffic in the city has been greatly challenged. Congestion has become synonymous with urban traffic, and the number of traffic accidents cannot be underestimated. However, with the rapid development of technologies such as big data, a large amount of traffic big data has been collected and studied. As an important part of promoting transportation development, traffic big data has the characteristics of huge data volume, multiple data types, and high real-time performance. Applying big data-related deep learning theories to analyze and mine the characteristics and trends of traffic big data will help improve the urban environment. The current situation of traffic congestion can not only solve th...

AI Technical Summary

Problems solved by technology

These works ignore the spatiotemporal association and task dependence between different prediction tasks, and only predict a single task, losing the task dependence between tasks, such as the association between departure time and travel time, different departure time Time corresponds to different travel times. If we can capture the dependencies between multi-tasks to better model traffic data, the prediction accuracy will be greatly improved.

Images