Grasshopper

algorithmic modeling for Rhino

# Evacuation simulation 01

This is a Rhino/Grasshopper definition. The goal is to simulate a tsunami evacuation scenario. This is a work in progress with fictional data, part of an ongoing research at the University de Concepción, Chile

Additionally to standard GH components I am using Spiderweb, Hoopsnake, Human, Elk and Ghowl.

Views: 252

Location: (Show Map)

### Related Videos

Comment

Join Grasshopper

Comment by Frane Zilic on June 23, 2013 at 2:26pm

Hi Richard

To approximate the initial density of each street section I start with the block density and I perform straight skeleton subdivision of the blocks at each side of the street to get a tributary area. I add this with singularities like schools, theaters, hospitals, etc... so I get a worst case scenario with the maximum amount of people in the street.

I guess I perform something similar to flow network, but since I am not that good at coding I made a geometrical solution. Each segment has its own shortest path to the secure point represented by a line somewhere in Z, and the density is represented by the height of a rectangle that travels along the line. (http://www.grasshopper3d.com/video/evacuation-simulation-02) The density is assumed to be homogeneous for each segment but not static as the people have different speeds so they spread and also respond to the slope of the path. At discrete points in the grid (the bars you see on the ground in this video) I evaluate the heights of all the rectangles above it, add them up and create the height of the bar. Then I calculate the density based on street width and that drives the colour.

I know this is not the most efficient or accurate way of proceeding but this is some sort of proof of concept that had to be developed in five days and that will be refined later.

Comment by Richard Schaffranek on June 20, 2013 at 7:15am

Hallo Frane,

yes I would be especially interested how you actually calculate the "human densities"?

Using some sort of flow network?

Comment by Frane Zilic on June 7, 2013 at 6:27pm

Hi Richard

Sorry I didn' t reply earlier.

This is a research collaboration between JICA and the ministry of infrastructure of Chile, to build tsunami-resilient communities. It is a very large project involving physical and social aspects, so everything from water simulation to the education of the community. In our group we must develop the evacuation plan and communicate it, and I am in charge of the evacuation simulation.

For now I am trying to understand the evacuation potential under the ideal situation to know what is theoretically possible.

First I identify the secure points in the city, protected from tsunami (the red dot in this case). Then I divide the city on cells,taking into account the area of influence of each secure point based on distance and capacity (I should consider time, capacity and access time but I'm not there yet). Next I analyze the grid and find the shortest paths to the nearest secure point. I sort all the paths and then I make "human densities" travel along those paths taking into account slope and  different speeds. Finally I get cumulative values that allow to evaluate the city and the access to the secure point.

I am still working with fictional data, there are some variables that are not considered yet, the code could be more efficient, but it's a start.

bye

Comment by Richard Schaffranek on June 3, 2013 at 1:15pm

Hallo,

Just found the simulation videos.

Is there more detail available to the research you are doing?

thx

Richard

• View All