Canopy rationalization

Daniel,

I have a geometry problem that I'm hoping Kangaroo can help me solve but I can't quite get the results I'm looking for.

I'm working on a canopy structure that is supposed to have a defined pattern in plan while being variable in the z direction so that it undulates like a tree canopy. Each panel needs a slope of 1/2" / foot for drainage (this is worked out in a separate file). 

The goal is now to reduce the number of unique glass panels for fabrication. My current strategy is to reduce the number of unique line lengths which will lead to a finite number of glass sizes. I've grouped similar line lengths and tried to use springs to pull them to their average, but the system is over constrained and resists resolving itself without drastically altering the geometry.

Any help you can offer would be much appreciated.