The basic goal of this research was to develop a parametric driver model that can be used to help evaluate and predict representative driver/vehicle handling behavior at or near the limit of driving performance. The model includes provisions for representing a range of driver steering characteristics and vehicle dynamic properties so as to permit computer-based investigation of different combinations of vehicle chassis properties and driver abilities beyond the so-called "normal" driving regime. The focus of this initial stage of work was on driver-vehicle steering interactions under near-emergency or surprise conditions that can require substantial utilization of available tire/road friction. Accommodation of speed changes and alteration of vehicle dynamic properties with lateral and longitudinal acceleration operating conditions--within the driver model--is included in the model. The nonlinear General Motors driver steering model developed under this work significantly extends analysis and predictive capabilities well into the nonlinear/near-limit handling regime of driver/vehicle systems. In addition to the extended nonlinear performance, the development work has also added certain new model features that include variable driver preview, driver speed control preferences based on upcoming road curvature changes, sensory input processing of vehicle response signals, path selection/adjustment options, and a simplified situational awareness element. (Author/publisher)
Abstract