Merging Strategy for Autonomous Vehicles on Highways Based on Acceptance Gaps and Model Predictive Control

Haiyang Zhao
DOI: https://doi.org/10.54097/1svzjz89
2024-06-26
Abstract:With the development of autonomous driving technology, highway ramp merging has become a critical challenge. The ramp merging area is a complex node in the highway system, where improper merging can lead to traffic congestion and accidents. Therefore, studying merging strategies for autonomous vehicles at highway entrances is of great significance for improving traffic safety and flow, as well as promoting the overall development of intelligent transportation systems. This paper proposes a merging strategy based on acceptance gaps and a Model Predictive Controller (MPC). A safety acceptance gap model and an experience acceptance gap model were constructed based on the Gipps model and historical data, respectively. By balancing merging efficiency and safety, a dynamic acceptance gap model with linear fusion was proposed. During the merging process, an MPC was used, which constructed the corresponding cost function and constraints based on the characteristics of highway ramp merging. Using the concept of dynamic programming, the MPC designed a merging trajectory controller. Finally, the effectiveness and feasibility of the merging strategy were verified through simulation experiments.
What problem does this paper attempt to address?