障害物の運動特性を考慮したポテンシャル場に基づく自律走行パーソナルモビリティの運動制御

抄録

In recent years, Personal Mobility (PM), which is a small, low-speed electric vehicle, has been attracting attention as a new means of transportation to replace walking and driving, and research and development of autonomous driving technology for PM, especially to support the disabled and elderly, are underway. Since PM has a high affinity with pedestrians, it is expected to be effectively used in “shared space,” which is a road space where pedestrians, bicycles, automobiles, and streetcars coexist. In order for PM to run autonomously in the “shared space”, it is important to have a motion control method that can safely avoid obstacles of various shapes, such as pedestrians and streetcars, while considering the comfort of the occupants. Conventionally proposed methods allow PM to smoothly avoid various obstacles based on the prediction of their motion, but if the prediction deviates significantly, it becomes difficult to avoid the obstacle. Therefore, in this paper, we propose a new motion control method for PM that takes into account the motion characteristics of obstacles. Specifically, we define a new potential field that reflects the kinematic characteristics such as the magnitude and frequency of directional and velocity changes caused by obstacles, and control the motion of the PM based on this field. This is an extension of the conventional potential field method as a path planning method. Various simulation results have shown the effectiveness of the proposed method.