Methodologies for evaluating the effectiveness of collision avoidance assistance system

Abstract

In this study, methodologies for estimating the accident mitigation ratio using the collision avoidance device are discussed. In these methodologies, the driver’s braking or steering and viewing behavior are investigated in a driving simulator. Braking or steering behaviour was identified in terms of the reaction time for braking or steering in order to avoid a collision or lane deviation and viewing behaviour was identified in terms of the duration from looking off a leading vehicle to looking back. Next, to investigate the combined effect based on the reliability of driver and the reliability of system, two different kinds of reliability-estimation model simulating its operating status were constructed. As an example study applying these methods, the effectiveness of the device supplying the aroma to increase the attention level for collision avoidance is quantitatively estimated through the statistical calculations based on the reliability-estimation model.

Keywords

• driving support
• collision mitigation
• total error of human-machine system
• monte-carlo simulations
• system reliability engineering