Modeling of Human Driving Behavior Based on Piecewise Linear Model

摘要

This paper presents development of the modeling strategy of the human driving behavior based on the expression as Piecewise Linear (PWL) model focusing on the driver's stopping maneuver. The driving data are collected by using the three-dimensional driving simulator based on CAVE Automatic Virtual Environment (CAVE), which provides stereoscopic immersive virtual environment. In our modeling, the control scenario of the human driver, that is, the mapping from the driver's sensory information to the operation of the driver such as acceleration, braking and steering, is expressed by Piecewise Linear (PWL) model. Since the PWL model includes both continuous behaviors given by polynomials and discrete logical conditions, it can be regarded as a class of Hybrid Dynamical System (HDS). The identification problem' for the PWL model is formulated as the Mixed Integer Linear Programming (MILP) by transforming the switching conditions into binary variables. From the obtained results, it is found that the driver appropriately switches the 'control law' according to the sensory information. These results enable us to capture not only the physical meaning of the driving skill, but also the decision-making aspect (switching conditions) in the driver's stopping maneuver.%Ovaj članak prikazuje razvoj strategije modeliranja ljudskog ponašanja pri vožnji, koja je utemeljena na po dijelovima linearnom (PWL) modelu fokusira-nom na vozačev manevar zaustavljanja. Podaci o vožnji prikupljeni su korištenjem trodimenzionalnog simulatora vožnje zasnovanog na CAVE Automatic Virtual Environment (CAVE) koji osigurava potpuno stereoskopsko virtualno okruženje. Pri modeliranju je upravljački scenarij za vozača, odnosno preslikavanje vozačevih senzorskih informacija u operacije poput ubrzanja, kočenja i upravljanja vozilom, opisan PWL modelom. Kako PWL model uklju-čuje istodobno kontinuirano ponašanje izraženo preko polinoma kao i diskretne logičke uvjete, takav se model može promatrati kao klasa hibridnih dinamičkih sustava (HDS). Transformiranjem uvjeta prekapčanja u binarne varijable, problem identifikacije PWL modela formuliran je kao mješoviti cjelobrojni linearni program (MILP). Iz dobivenih je rezultata vidljivo da vozač prekapča »zakon upravljanja« u skladu sa senzorskim informacijama. Rezultati omogućuju razumijevanje ne samo fizikalnog značenja sposobnosti vožnje već i sam aspekt donošenja odluka (uvjeta prekapčanja) prilikom vozačevog manevra zaustavljanja.