Page 229 - Big Data Analytics for Connected Vehicles and Smart Cities

210 Big Data Analytics for Connected Vehicles and Smart Cities Practical Applications and Concepts for Transportation Data Analytics 211

In Figure 10.8, two additional boxes have been added—the leftmost black
boxes that are labeled “TMC LAG1” and “TMC LAG2.” These represent post-
bottleneck traffic speed conditions.
This post-bottleneck traffic zone can also be defined as the zone of influ-
ence of the bottleneck, and the zone speed has been defined as the average of
speeds observed in TMC LAG1 and TMC LAG2. The addition of these two
segments makes it possible to conduct analysis on the zone of influence with
respect to traffic turbulence.
To provide an intuitive measure of the impact of turbulent speed change
between adjacent TMCs, we adopted a calculation of acceleration/deceleration
and cumulatively calculated the total absolute value of these acceleration/decel-
eration calculations. The methodology for measuring the acceleration between
two adjacent TMC’s (x and y) is as follows:

sum across all pairwise TMC segments of 
 
x
 TMC _ _speed ∗   Total number of 

  
x
Traffic turbulence =   TMC _ _speed    
  TMC _ _speed ∗  pairwiseTMC segment 
y
−    comparisons 
 TMC _ _speed    
y

 
where
TMC x speed is the speed observed on segment X for 1 minute;
TMC y speed is the speed observed on segment Y for one minute;
TMC x pmm is the pseudomile marker for segment X;
TMC y pmm is the pseudomile marker for segment Y.

The measurement units for the above calculation are (miles/hour) per
minute
Developing this analytic for peak period traffic on weekdays, using the
two post-bottleneck TMC segments, TMC LAG1 and TMC LAG2, yielded
the results summarized in Figure 10.9.
Figure 10.9 shows the results of the turbulence analysis for the fourth
quarter of 2014 compared to the fourth quarter 2013. As explained earlier the
VSL signs were turned at the beginning of the fourth quarter of 2014, defining
the after period. The results show a clear reduction in turbulence and speed at
the end of the bottleneck. The result was not checked for statistical significance
but was of sufficient significance to satisfy the client’s needs. Note that there are
approximately 71.5 million data points in the turbulence calculation.