A Novel Scenario-based Path Planning Method for Narrow Parking Space

Automatic parking is an important field in autonomous driving research, and path planning is a crucial step in its realization. However, traditional path planning algorithms tend to fail when confronted with narrow parking space. In this paper, a novel scenario-based path planning Method is proposed for narrow parking space. First, the scenario decision algorithm is used to determine the parking scenario category: parallel parking or perpendicular parking. Then suitable intermediate state in configuration space is selected through the backtracking method and the local path of intermediate-goal is generated. After that, the initial-intermediate path planning is performed. Finally, the global path is obtained by concatenating the two local paths. Experiments are conducted to evaluate the performance of the proposed method and the results show that it can not only quickly generate a path that satisfies vehicle kinematic constraints in narrow parking space.