A driver-distraction-based lane-keeping assistance system

Abstract

Single-vehicle roadway departure (SVRD) accidents occur in many cases owing to driver distraction or drowsiness constituting a substantial share of today's road vehicle accidents and casualties. This paper describes a distraction-based lane-keeping support system, which intervenes only when the driver is positively detected as being distracted. Distraction here is understood as cognitive and visual distraction, and the focus of this system is on the latter one. In order to estimate the driver's visual distraction level, a video-based driver monitoring system is used. Lane-keeping support is provided by an additional torque applied on the steering shaft in order to regain an appropriate lane position. In this manner the system only intervenes when the vehicle has drifted out of its lane and while the driver is distracted. Test track investigations indicate large opportunities for such a system from a driver perspective, provided that sufficient reliability of the employed vision sensor for lane and face tracking can be obtained.

Driver inattention monitoring system for intelligent vehicles: A review

“…But to model the link between eye movement and visual & cognitive distraction is quite difficult, especially for the latter since the location of our fixation does not always reveal the target of our attention. J Pohl et al [50] modeled the visual distraction level as a time dependency of the visual focus, with the assumption that the visual distraction level is non-linear: visual distraction increases with time but decreases nearly instantaneously. They established their algorithm for visual distraction: first there is a distraction calculation to compute the instantaneous distraction level, second there is a distraction decision-maker to determine if the current distraction level represents a potentially distracted driver.…”

A unified approach to semi-autonomous control of passenger vehicles in hazard avoidance scenarios

“…The vehicle model described by (2)(3)(4)(5), with the parameters given in Table 1 was used in the receding horizon controller. Controller parameters are defined and quantified in Table 2.…”

“…Among existing proposals for semi-autonomous vehicle navigation, lane-keeping systems using audible warnings [4], haptic alerts [5], steering torque overlays [6], and various combinations of these have been developed with mixed results [7]. In a recent subproject of the European PReVENT consortium, a lane-keeping system was designed to prevent lane departure by perceiving the environment, making heuristic-based trajectory planning decisions based on perceived threat, and implementing warning mechanisms or slight steering torque overlay when the vehicle drifts from the desired trajectory [8].…”

Multi-objective decoupling algorithm for active distance control of intelligent hybrid electric vehicle

“…To reduce fuel consumption and exhaust emission, a variety of clean energy vehicle technologies have been widely studied: pure electric vehicles [1], hybrid electric vehicles [2], plug-in hybrid vehicles [3], fuel cell vehicles [4] and alternative fuel vehicles [5]. In-depth study to improve safety performance of ICE vehicles has focused on active safety technology and intelligent safety technology, such as Anti-Brake System (ABS) [6], Electronic Stability Program (ESP) [7], Global Chassis Control (GCC) [8], Adaptive Cruise Control (ACC) [9], and Lane Keeping System (LKS) [10]. Integration of active safety technologies in clean energy vehicles have become increasingly important [11], as illustrated by the application of the ABS [12], TCS [13] and ESC [14] in pure electric or hybrid electric vehicles.…”