Estimation of Steering Model depending on Gazing Distance in a Curvy Road

Abstract

To analyze the driving behavior, the steering model is important for designing a comfortable automatic driving. It is known that the PID gain of the model is assumed to be modified depending on driver's gazing distance. The aim of this study is to design a steering model with the gazing distance using Particle Swarm Optimization (PSO). The steering model was estimated well in a single curve, however, the steering model in a curvy road have not estimated yet, because it is expected the drivers changes according the situation. In this paper, the estimated method of the steering model is considered in a curvy road. For this, the switching of a hybrid model based on effective visual field is proposed, that consists of PD control and on-off control models. The effectiveness of proposed method is evaluated by simulation using Honda driving simulator.