CHAPTER 6. CONCLUSION
70
|
controller |
density |
ATWT |
TWQL |
|
marching |
HD |
264 |
1434 |
|
optim |
HD |
235 |
1790 |
|
request |
HD |
344 |
1457 |
|
phase |
HD |
89 |
144 |
|
platoon |
HD |
89 |
155 |
|
marching |
LD |
175 |
428 |
|
optim |
LD |
216 |
481 |
|
request |
LD |
31 |
0 |
|
phase |
LD |
38 |
0 |
|
platoon |
LD |
18 |
0 |
Table 6.1: Scenario 1: Best results
From table 6.1 can be concluded that in almost all cases the self-organizing
traffic lights are better than the marching and optim controller. The sotl-request
controller for high traffic density is worse. This is so because at high traffic density
requests are granted almost immediately. This results in the very fast switching of
traffic lights. The value of θ can regulate the green times of the traffic lights. This
is why the highest value for θ gives the best ATWT value (Average Trip Waiting
Time).
The sotl-phase controller is better than the optim and marching controller for
both traffic densities. Compared to the optim controller, the sotl-phase controller
at high traffic density reduces the ATWT with 62% and the TWQL (Total waiting
queue length) with 91%. At low traffic density, this is a reduction of 82% for the
value of ATWT and an elimination of the TWQL value.
The sotl-platoon controller has very good results. Compared to the optim
controller, the sotl-platoon controller has reduced the ATWT with 62% and the
TWQL with 90% at high traffic density. At low traffic density, the sotl-platoon
controller reduces the ATWT with 90% compared to the optim controler and the
TWQL is eliminated.
6.2.2 Scenario 2
Scenario 2 is an infrastructure with five junctions, 12 edge-nodes, and 16 roads in
two directions. There is a bigger traffic flow from edge-node 0 to edge-node 11