Hi everyone,
I've recently found myself in the unusual position of having some extra time on my hands and I've devoted it to a project that's been simmering for a while on the back burner: creating a building energy optimization algorithm using DOE-2 as a simulation engine. I've begged, borrowed, stolen, and tinkered my way to a Python script that can edit .inp files, execute DOE-2 from the command line, and process .sim file text to return a value for an objective function (for the sake of an example, let's say annual energy-related carbon emissions). My next step is hooking this up to an appropriate optimization algorithm. For the moment, I'm dealing with simple problems with a handful of independent variables like orientation, U-values and SHGCs, window to wall ratios, etc.
Are there any suggestions for an appropriate algorithm (and hopefully a software implementation) that handles these problems well? I recall the eQuest team's impressive optimization exercise in the Lowdown Showdown a couple years back, so I'm curious about any of the resources they used. Was the algorithm implementation completely homebrewed, or did it use any off-the-shelf components?
Thanks, Drew
