A multi-lane space-time trajectory optimization method for intelligent networked vehicles

Abstract

The invention provides a multi-lane spatio-temporal trajectory optimization method for intelligent networked vehicles. The invention designs a spatio-temporal trajectory optimization algorithm based on a reinforcement learning algorithm to quickly match the optimal trajectory. The algorithm includes the calculation of two different inputs: (1) the optimization of the space-time trajectory, taking the vehicle's current position, speed, target exit lane, and time period as the input, and the set of vehicle acceleration as the output; (2) multi-lane cooperative change The optimization of the lane takes the current position and speed of the vehicle and the position and speed of the threat vehicle in the target lane as input, and the vehicle acceleration set as the output. That is, after the vehicle initiates a lane change request, reinforcement learning can be used to match the trajectory of the vehicle's collaborative lane change process. After the lane change is completed, reinforcement learning can be used to match the spatiotemporal trajectory at this moment to achieve the process of multi-lane trajectory optimization. This method can optimize and generate the spatio-temporal trajectories of passing vehicles in a road section in real time according to different road environments and traffic conditions, which increases the mutual cooperation ability between vehicles, improves the safety of passing through road sections and the efficiency of vehicles passing at intersections, and reduces the traffic flow of vehicles. In order to ensure road traffic safety and improve travel efficiency, new solutions and theoretical basis are proposed.

Description

technical field [0001] The invention belongs to the technical field of vehicle-road coordination / arterial traffic flow control, and specifically relates to a multi-lane space-time trajectory optimization method for intelligent networked vehicles, which is applicable to any signalized intersection section in an urban road traffic network. Background technique [0002] For the urban traffic road network, the current urban traffic road system controls the traffic flow at the intersection, except for a small number of suburban intersections with small traffic volumes that adopt the self-organizing control method without signals. The intersection is the main node connecting each road section, and the reasonable planning of traffic flow in each road section is also an important part of improving the traffic efficiency of the intersection. The driving behavior of the vehicle during the driving process of the road section can be divided into lane keeping behavior and lane changing b...

AI Technical Summary

Problems solved by technology

Compared with the single-lane space-time trajectory optimization problem, the multi-lane space-time trajectory optimization problem has higher complexity, and it is difficult to directly solve the multi-lane space-time trajectory optimization problem through the existing lane changing rules

[0013] 2. Traditional routing methods have limited information acquisition capabilities, often rely on fixed traffic detectors, and lack real-time, efficient, and accurate vehicle driving information and traffic status information

[0014] 3. Most of the traditional routing methods do not consider the factor of multi-vehicle coordinated lane change, and lack of consideration of the interference factors of surrounding vehicles

Establishing a mathematical model for the existing lane-changing rules is difficult to meet the multi-vehicle goal at the same time; for the mathematical calculation of the system, it is difficult to process a large amount of data of multiple vehicles at the same time

Images