aufdenkampe
commented
2 years ago From @ptomasula's 2021-08-20 "HSP2: GENER Module LOE" email:
Supporting the GENER functionality as implemented by our use case test is notably more complicated than just writing the GENER module (which in itself is actually relatively straight forward).
Here’s a rough outline of the steps that need to be taken:
- Update the UCI reader:
- Modify function that processes OPSEQ block to import both the COPY and GENER operations.
- Also import the GENER and COPY block, which necessitates adding new tables to the HDF5 files and writing new functions in the UCI reader.
- Modify functions that process NETWORK and SEQUENCE blocks to import COPY and GENER related links.
- Modify function that processes EXT SOURCES block to import COPY related lines.
- Write COPY module, which should essentially be a class that reads timeseries as specified in EXT SOURCES and holds those timeseries in memory. There’s also the option to apply scaling factors, but we can use the main.read_timeseries function as a template.
- Write GENER module, which is also a class that should be able to hold the timeseries, post calculation, in memory.
- Add new dictionaries to main.py to store and track COPY and GENER class instances.
- Add new function that can be called at beginning of RCHRES, PERLAN, and IMPLND to fetch data from the appropriate COPY/GENER instances where applicable. This is similar to the existing get_flows function in main.py but with some notable differences that warrant this being a separate function.
The GENER module is used extensively to compute (within the model) constituent loads based on input flow and concentration, with one or both being a time series.
These capabilities are used extensively several of our HSP2 use cases
@ptomasula is getting started on this.