Multilayer Modeling of Driver Behavior Based on Hierarchical Mode Segmentation - Digital Signal Processing for In-Vehicle Systems and Safety - page 60

1

0.5

0

1

0

-1

1

0

-1

1

0

-1

1

0

-1

1

0.5

0

1

0.5

0

1

0

-1

1

0

-1

200 300 400 500 600 700 800

Time[sec]

Fig. 4.4 Observed profiles and mode segmentation result (Examinee A, two modes)

Figure 4.3 shows that when the two clusters are unified, the corresponding
dissimilarity is designated by the horizontal bar. The horizontal axis represents
the data which has been rearranged after clustering to indicate the hierarchical
structure. From this figure, we can clearly understand the hierarchical structure of
driving behavior. Two dashed horizontal lines are superimposed. The upper line
indicates the number of modes (clusters) s , i.e., the number of the ARXmodels in [ 2 ]
is set at two. Meanwhile, the lower line shows that s is set at five.
In Figs. 4.4 and 4.5 , the observed driving (input-output) profiles of examinee
A are shown. All profiles are normalized before clustering. In the profile of the
lateral displacement, it takes positive value when the examinee's vehicle is on
the right side of the lead car. The steering angle takes positive value when the
examinee steers it clockwise. Also, the pedal operation takes positive value when
the accelerator is stepped on, and takes negative value when the examinee hits
the braking pedal. Note that the range, range rate, and lateral displacement
profiles demonstrate discontinuity. Since these variables are defined by relative