A Game Lane Changing Model Considering Driver’s Risk Level in Ramp Merging Scenario

A ramp merging decision as an important part of the lane change model plays a crucial role in efficiency and safety entire process. However, due to inevitability on-ramp merging, limitations road environment, conflict between vehicle following on main road, it is difficult for human drivers make optimal decisions complex scenarios. First, based NGSIM dataset, gain function designed represent interaction ego (EV) surrounding vehicles, value then used one characteristic parameters. The K-means algorithm employed conduct cluster analysis driving style under condition changing lanes. This paper models (vehicle merging) mainline lagging complete information non-cooperative game Further, various styles are coupled mimic different travel preferences drivers. After EV decision-making, quintic polynomial method with multi-constraints proposed implement trajectory planning. tested analyzed scenario, results demonstrate that can correct merging. trend speed tests also line features driver’s style, offering theoretical foundation individualized decisions.