Dynamic Carpooling Recommendation Method for Commuter Private Cars

Abstract

The invention discloses a dynamic carpooling recommendation method for commuting private cars, which includes inputting automatic license plate recognition data, preprocessing it, and generating a vehicle trajectory set; based on the urban vehicle trajectory set, combined with the high repeatability of the spatiotemporal rules of commuting private cars and periodicity, design a workday commuting private car identification algorithm; combine the home-work address, commuting time window and workday dynamic factors (weather, accidents and traffic) of commuting private cars to realize the impact of commuting private car carpooling Key feature extraction; based on the spatio-temporal characteristics of commuting vehicles, matching commuting vehicles with similar commuting spatio-temporal patterns, and combining the impact of dynamic factors on drivers' willingness to carpool, effectively combining these dynamic factors with users' carpooling needs in the spatio-temporal dimension to achieve Dynamic carpooling matching based on user wishes provides a long-term and stable carpooling mechanism for commuting private cars, and significantly reduces the number of private cars during morning and evening peak hours.

Description

technical field [0001] The invention relates to the field of intelligent transportation systems, in particular to a dynamic carpooling recommendation method for commuter private cars. Background technique [0002] With the rapid development of location-based services and mobile Internet technologies, carpooling services have become a new sharing economy model for alleviating urban traffic congestion. [0003] In the field of carpooling recommendation, there are two problems with traditional algorithms: First, most of the existing methods have a single goal of meeting the real-time carpooling needs of passengers, that is, how to reduce the waiting time of passengers. And the problem of increasing the use of private cars. Therefore, existing carpooling methods have the potential to cause more severe road congestion during morning and evening rush hours. Second, most of the existing carpooling schemes rely on large-scale crowd movement data such as GPS trajectories and mobile...

AI Technical Summary

Problems solved by technology

[0003] In the field of carpooling recommendation, there are two problems with traditional algorithms: First, most of the existing methods have a single goal of meeting passengers’ real-time carpooling needs, that is, how to reduce passengers’ waiting time, and have not effectively solved the problem of competition among carpooling drivers. And the problem of increasing the use of private cars

As a result, existing carpooling methods have the potential to cause more severe road congestion during morning and evening peak hours

Second, most of the existing carpooling schemes rely on large-scale crowd movement data such as GPS trajectories and mobile phone signaling. The trajectories are rich but semantic information is lacking, and the influence of dynamic factors such as weather, holidays, major events, and traffic accidents on the carpooling scheme has not been effectively considered. Significant influence

Therefore, it is difficult to provide users with personalized dynamic carpooling solutions

Images