Automatic simulating EnergyPlus with different occupancy schedule

edit

Hey Brian, thanks for your answer! I am wondering if i am using openStudio server, should i re-create all HVAC system in detail? Coz, what I have now is quite complex which I really dont want to create new hvac system in OpenStudio again.

The occupancy schedules are generated from my other modelling, when you say 'You could write a measure to apply your schedule changes to the models and then run the analysis on Amazon EC2' Is this including all schedules or I should change the schedules manually? Thanks

You can still use OS-server with your original IDF. Set the seed model to be the EmptySeedModel.osm that is in the /seed directory. Then use the InjectOsmGeometryIntoAnExternalIdf measure that is in the /measures directory and give it the path to your IDF and set the Merge Geometry argument to FALSE. This will essentially skip all the OS functionality and insert your IDF into the workflow. From this point on, you can use an E+ measure to change/replace the schedules you want. The trick will be to parameterize the measure with an argument that will change the schedules the way you want.

Once you parameterize the measure with the argument that changes the schedules the way you want, you can turn that measure argument into a variable in the OS-spreadsheet and then apply an algorithm (like sampling, or full factorial design) to run all the combinations you are looking for.

An example would be to make a 'choice' argument that is the name of the schedule you want to add to the model. Write the measure so that it takes that 'choice' name, and applies the schedule changes correctly to the model. Then you can give a discrete list of choice values to the spreadsheet to run.

@nancynanxiaodan the most straight forward approach for changing out schedules would be to pre-load the possible schedules into your IDF file. The the measure will just swap out schedule assignments based on user arguments. You could keep the schedules in separate IDF files or generate them on the fly from a CSV file, but unless there is a reason not to, just put them in your original IDF.

I don't know if we have an EnergyPlus measure that does exactly what you want, but I can try to put some sample code together in another comment later this afternoon.

arg_string would be user argument. "orig" would be contained in name of schedules in seed idf.

# change people schedules
people = workspace.getObjectsByType("People".to_IddObjectType)
people.each do |object|
  orig_sch = object.getString(2).get
  object.setString(2,orig_sch.gsub('orig',arg_string))
end
# change heating schedules
thermostats = workspace.getObjectsByType("ThermostatSetpoint:DualSetpoint".to_IddObjectType)
thermostats.each do |object|
  orig_sch = object.getString(1)
  object.setString(1,orig_sch.gsub('orig',arg_string))
end

Dear Jamie, thanks for your help here. I've looked at the Eppy which seems pretty easy to understand its capability from the web. From my understanding that what Eppy's capability is that it could easily make change of the IDF file. May I ask that if all the occupancy schedules are loaded in the IDF file, could EnergyPlus simulate them all at the same time? Or what it will do is change schedule and run the simulation and then change the schedule again, and then run the simulation again, and repeat the process to end. And about the Raven at my university, could be used for it. Thanks very much.