Original
Optimal
Assorted Node Logic
Raw Data Exported from Custom Tower
To optimize the building choice, I wanted to consider a cost vs. value objective. To do this, I recreated the linear cost module from the example and then made another similarly structured node that instead scaled quadratically from 500$ to a bit over $2000 over the 750 ft. My thinking was that the super high rise real estate would increase more than linearly in value, but not cost of construction, thus giving motivation to have more square footage at elevation. I then divided value by cost and normalized these results to create a proxy for real estate value. I then used the solar insolation function and divided the average insolation by the surface area. This should provide a proxy for the cost of heating. I then normalized this data as well. To optimize a single objective function, I multiplied one set of normalized data by the other and chose the options with the highest value. The optimal building in my scope of parameter choices was the pentagonal hourglass tower with 750’ height and 50’ middle radius. The runners up were the twisting triangular tower with 750’ height and 90-degree top rotation and the pentagonal hourglass tower with 750’ height and 60’ middle radius (the triangular tower’s midpoint rotation was manipulated to be half the top rotation so the toward would have a constant twist.