Forward Collision Probability Evaluation Based on Tired Index of Driver

Abstract

-This paper presents a forward collision probability evaluation based on tired index of driver for alerting and assisting driver to avoid the forward collision in highway driving. From electroencephalograph (EEG), we could figure out the relaxed index (RI) and focus index (FI) of the driver. And then a tired index (TI) of driver will be found. We use the tired index of driver, the time-to-collision (TTC) and headway time (HT) to be the factors for evaluating the forward collision probability. When we evaluate the forward collision probability indices of TTC and HT, a Mamdani fuzzy inference algorithm is presented for calculating the forward collision probability index of the vehicle. The forward collision probability index (FCPI) is easily understood for the driver even who does not have any professional knowledge in vehicle field. At the same time, the behaviour of driver is taken into account, called the selflearning algorithm, for each of driver using. For the FCPI, the value 0 indicates the 0% probability of forward collision, and the values 0.5 and 1 indicate the 50% and 100% probabilities of forward collision respectively. It is useful and easily understood for the presented probability index of forward collision for alerting driver to avoid the forward collision when driving in highway.