Target Space Selection for Automatic Lane-Changing System at Congested Highway On-Ramp

In this study, we construct a method for selecting a target space when changing lanes from the merging lane to the main lane, assuming a merging scene on a congested highway. The proposed method predicts the driving trajectory for a few seconds ahead based on the state variables of the ego vehicle and surrounding vehicles, which is used to evaluate the lane change target. We determine the reachable range of the ego vehicle and select a candidate group of inter-vehicle spaces that will serve as the target for the lane change. Next, the proposed method evaluates the candidate group using indicators such as the size of the inter-vehicle space, the distance from the ego vehicle, and the remaining distance of the merging lane, selecting the inter-vehicle space with the highest evaluation as the target for the lane change. Through simulation experiments where the diversity of driving characteristics of human drivers on the main lane is considered, we confirmed that the proposed method has sufficient safety and stability.