Automatic driving multi-vehicle intelligent cooperation regional traffic flow guiding method

Abstract

The invention discloses an automatic driving multi-vehicle intelligent cooperation regional traffic guidance method, which is applied to the field of intelligent Internet of Vehicles, and aims at the problems that a traditional path planning algorithm is short in sight and ignores communication resource constraints. The invention provides a multiple-inteligent-agent depth deterministic strategy gradient model with a self-evolution function, and the model can carry out centralized training and distributed execution on road network data, that is, each automatic driving vehicle can have different reward structures and make decisions according to local information of the automatic driving vehicle. Strategy congestion caused by strategy conflicts is avoided to a certain extent; and an area server establishes digital twinning for vehicles in the area to perform road condition deduction, and generates periodic feedback for the vehicles. According to the invention, through a digital twin iteration vehicle strategy, future traffic conditions under the action of multiple vehicles are accurately speculated and are fed back as input to the vehicles for distributed decision making. On the premise of considering personal travel safety and time constraint, the multi-vehicle travel efficiency in the system is remarkably improved.

Description

technical field [0001] The invention belongs to the field of intelligent vehicle networking, and in particular relates to a multi-vehicle intelligent collaborative area traffic flow guidance technology. Background technique [0002] Effective route planning plays a key role in improving transportation efficiency. Without comprehensive real-time traffic information, drivers can only make routing decisions based on their own limited field of vision. These short-term and non-coordinated routing choices will inevitably cause a large number of conflicts on planned routes and reduce the resource utilization efficiency of the road network. In recent years, the continuous improvement of vehicle sensing, computing and communication capabilities has provided opportunities to improve urban transportation. Specifically, the 5G-enabled vehicular ad hoc network (5G-VANET) facilitates the exchange of real-time traffic information between vehicles and infrastructure. In addition, cloud an...

AI Technical Summary

Problems solved by technology

[0007] In order to solve the short-sightedness of traditional traffic flow control and its ignorance of communication resource constraints, the present invention provides a traffic flow guidance method based on traffic and communication resource perception for autonomous multi-vehicle intelligent collaborative areas. The multi-agent reinforcement learning technology of the model establishes a digital twin of the self-driving vehicle in the regional server, calculates the road conditions in the future, and feeds back the results to the vehicle for its distributed decision-making, and adds communication resources to the traditional The road network realizes the perception and decision-making of autonomous vehicles on future traffic conditions in the region, and improves the efficiency of multi-vehicle travel in the system on the premise of meeting individual needs

Images