2P1-D07 Stochastic Switched ARXモデルを用いたドライバの運転行動認識

概要

論文の詳細を見る
This paper presents a development of modeling and recognition of the human driving behavior based on the expression as Stochastic Switched ARX model (SS-ARX). First of all, the parameter estimation algorithm for the SS-ARX model is developed based on the EM algorithm. This can be achieved by extending the parameter estimation technique for conventional Hidden Markov Model (HMM). Second, the parameter estimation algorithm is applied to the driving data focusing on the drivers' collision avoidance behavior. The driving data were collected by using the driving simulator based on the stereoscopic immersive VR system CAVE. Then, the parameter set for each driver is found and some driving characteristics can be understood. Finally, the performance of the SS-ARX model in the case of using as the behavior recognizer is examined. The results show the high potential ability of the SS-ARX model as the behavior recognizer.
一般社団法人日本機械学会の論文