Systematical Trajectory Planning for Cooperative Merging at Multilane On-Ramp Intersection

Traffic stream merging at on-ramp intersections often leads to tremendous traffic accidents, delays, oscillations, and pollution. Although various control approaches have been developed to address this problem, state of art shows that we still lack efficient methods to systematically improve traffic efficiency in this area. To partially make up the research gap, this study developed a systematical optimal trajectory planning (SOTP) scheme to promote cooperative driving (car-following and merging maneuvers) among CAVs in a multilane mainline-ramp intersection. The SOTP scheme was mathematically modeled as a mixed integer nonlinear program (SOTP-MINLP) to maximize stream traffic speed subject to vehicle dynamics and safety constraints. Moreover, we designed a novel parallel computing aided solution algorithm, which integrates the problem’s unique features into the decomposition, relaxation, and constraint enforcement techniques, so that we can solve the SOTP-MINLP efficiently and implement our approach for real-time traffic management. Our numerical experiments built on field data confirmed the applicability and functionality of the SOTP and its solution approach for improving traffic efficiency at a multilane mainline-ramp intersection.