A Study of the Jerk Cost Function for Evaluating Quantitative Driving Performance during Simulated Driving
The purpose of this study was to test whether the jerk, the third derivative of the position, can be used as a quantitative driving performance index during simulated driving. Sixty healthy adults having 1-3 years driving experience were participated in driving simulator experiment. All subjects were instructed to keep constant distance from front vehicle running at 55~65km/hr speed on straight road. Each 10 males and 10 females were randomly selected in condition of keeping a constant distance with 20m, 25m and 30m, respectively. Experiment consisted of straight driving for 1 min and unexpected situation caused by emergency stop of the front vehicle. To collect 3D kinematic data, 10 markers were attached on the subjects’ upper (elbow, wrist and hand) and lower (knee, exterior knee, ankle, and toe) limbs and 3D motion capture system was synchronized with the driving simulator. The data was divided into unexpected situation section for 2 sec. To evaluate the driving performance, the jerk cost function (JCF) which indicates the smoothness of motion was calculated for all makers. The probability of collision was significantly smaller at the distance of 30m compared with 20m and 25m (p<.05). JCF of upper limbs did not show any significant patterns in distance between vehicles for unexpected section. However, JCF of lower limbs (knee, ankle and toe) showed significantly smaller values as the distance between vehicles was increased for unexpected situation section (p<.05). The probability of collision was decreased and the driving performance was increased with decreased JCF as the distance between vehicles was increased. Thus it can be concluded that JCF could be used as one of the parameters for evaluating the driving performance quantitatively.
KeywordsDriving performance Jerk-cost function Simulator Motion analysis Unexpected situation
Unable to display preview. Download preview PDF.