Jordan Hurley
Environment Modeling, Adaptation
Need a proposal
NetLogo Traffic modeling
Oct. 3 - Looking at Conway's life variations
Oct. 6 - Looking at the NetLogo modeling of Traffic Two Lanes.
Running with look-ahead at 1 gives a higher average speed.
Oct. 11 - Using Netlogo practice programming turtles, computer froze
Oct. 13 -
Oct. 18 - Current applet version of traffic scenario - press setup, then go or go once.
Cars "wrap around", total 20 cars right now (but this value can be changed)
Oct. 20 - Working on a version that won't wrap around, i.e. new cars will appear on the left and travel across the image.
The newer models will be named current.html" at some point.
Nov. 17
Dec. 20 - Current Status Update
Progress as of December
As of now, I have almost completed lane changing. Once that happens, the
behavior of the cars in the simulation will much more closely model that of real
life. The main problem for me with that right now is setting an appropriate
tolerance or patience, so they switch lanes a realistic amount. If I set in
wrong, there will be hardly any difference, or on the other side, they will
switch back and forth frantically. Another issue is trying to prevent them from
all wanting to be in one lane.
The next thing I will do is not as coding-intensive. Right now, 2/3 of the
space used to display the simulation is wasted space, just green area that the
cars can not travel on. Once I make the display much wider than it is tall, that
allows for more behavior to emerge in the cars in the simulation, since they
have many more steps in which to stay on the screen.
The last thing is trying to keep track of memory. Currently, I have way too
many variables for each car. Every time one car takes a step, there are multiple
things that must be checked before it acts. I must streamline the coding if I
want to be able to scale up the simulation to be able to include many many cars.