Monday, May 2, 2016

Traffic Modeling

Last week, I became momentarily frustrated with a particular research problem I've been grappling with for over a year and a half now. Fearlessly oversimplifying, I found that some objects moving in crowded environments slow down below their natural speed as the surrounding objects become fast enough.

The literature and models in my sub-sub-subfield weren't helping. So I thought I'd look at research in traffic modeling. From personal anecdote driving on interstate highways, I've noticed that when there are a few rash drivers swerving lanes and driving superfast, I tend to be more careful and slow down.

I didn't find exactly what I was looking for, but bumped into a bunch of "popular" resources, which I thought I'd archive for a closer look later.

1. Here is a provocative, readable but slightly flawed analysis on the physics of traffic jams
2. The principal claims of the article are contested by a ridiculously simple model here.
3. Wikipedia has an interesting overview of traffic simulation
4. A slightly more research oriented (but very readable) resource at MIT with plenty of links

The first two articles provide radically different ideas on how to treat traffic jerks. The first article (and some other published research) argue that a few rule-breaking mavericks help. The second link suggests that due to physical constraints anybody who cuts into the "line" ahead of you, delays your estimated time of arrival.

No comments: