Abstract
It is important to clarify the driver’s avoidance behavior regarding irregularly moving obstacles such as pedestrians, bicycles and so on. In driver support systems or the autonomous driving vehicles, it is considered that the driver model using risk potential is effective. However, it is not considering the dynamic characteristics of the object at obstacle avoidance. We have already clarified the control of the driver at obstacle avoidance and constructed the driver model using the risk potential. We remodeled into the driver model that can be calculated in real time. This driver model could describe the driving behavior for the static obstacle avoidance. However, at the moving obstacle avoidance, the driver model could only describe the driver to a certain level. It is necessary to clarify the driving behavior and describe in more detail. Therefore, we constructed the risk potential algorithm considering the dynamics or state of obstacles using Kalman filter. As the result, the constructed driver model expresses the driving behavior, and we proposed a new control algorithm using the risk potential considering the dynamic of obstacles.