Inspiration

Last weekend, I and my friends went to Yilan for participating in activities. For God's sake, we were sitting in a traffic jam for two hours. When we finally got there, the show was nearing completion, that was really awful experience. Next day morning, I used google search to find out if there is anyone to solve the traffic jam problem, I found there are so many ways to this issue, but the problem still exist. It turns out I want to find my way to solve it.

What it does

To predict the future state of Hsuehshan Tunnel traffic.

How I built it

We read XML data by using python and analysis it.

method 1:

My entry point is traffic flow. How to quantify Traffic flow state? We use traffic clustering model to identify the basis of traffic flow state. For making calculation and predictive accuracy, we established an optimization model to predict traffic flow parameters which for regional traffic flow based on temporal-spatial-historical correlation. Finally a two-stage model is choosing to simplify the optimization model.

method 2:

Utilize the section speed, the section density and the section traffic volume, divide the traffic jam into six grades according to the Kmeans method, label it, and use the SVM to predict the section speed in the next section.