A method for detecting drivers' intentions is essential to facilitate operating mode transitions between driver and driver assistance systems. We propose a driver behavior recognition method using Hidden Markov Models (HMMs) to characterize and detect driving maneuvers and place it in the framework of a cognitive model of human behavior. HMM-based steering behavior models for emergency and normal lane changes as well as for lane keeping were developed using a moving base driving simulator. Analysis of these models after training and recognition tests showed that driver behavior modeling and recognition of different types of lane changes is possible using HMMs.

Longitudinal vehicle control and/or warning technologies that operate in accordance with drivers' subjective perception of risk need to be developed for driver-support systems, if such systems are to be used fully to achieve safer, more comfortable driving. In order to accomplish this goal, it is necessary to identify the visual cues utilized by drivers in their perception of risk when closing on the vehicle ahead in a car-following situation. It is also necessary to quantify the relation between the physical parameters defining the spatial relationship to the vehicle ahead and psychological metrics with regard to the risk perceived by the driver. This paper presents the results of an empirical study on quantification and formulization of drivers' subjective perception of risk based on experiments performed with a fixed-base driving simulator at the Nissan Research Center. Experiments were carried out to investigate the subjective perception of risk relative to the headway distance and closing velocity to the vehicle ahead using the magnitude estimation method. The experimental results showed that drivers' perception of risk was strongly affected by two variables: time headway, i.e., the distance to the lead vehicle divided by the following vehicle's velocity, and time to collision, i.e., the distance to the lead vehicle divided by relative velocity. It was also found that an equation for estimating drivers' perception of risk can be formulated as the summation of the time headway inverse and the time to collision inverse and that this expression can be applied to various approaching situations. Furthermore, the validity of this equation was examined based on real-world driver behavior data measured with an instrumented vehicle.

By providing robust real-time detection of driver lane changes, the system shows good promise for incorporation into the next generation of intelligent transportation systems.

No abstract

No abstract