ryancoe
commented
3 years ago @nickross4444 - Good question.
Why are we using a JONSWAP on the WEC controller?: Yes, this is a bit counter-intuitive. The JONSWAP is a representation of real ocean waves, so it makes sense that we would use this to drive the wave maker. The purpose for using this mode on the WEC is different. In this case, we are using the spectrum to do system identification (SID) on the WEC. The process is something like the following:
- Excite the system (in this case the WEC) with broad-banded signal. Ideally, the broad-banded signal would be white noise. However, we figured to save you time we could use the JONSWAP spectrum.
- Post-process the results: basically compare the input (in this case the JONSWAP) to the output (e.g., the velocity of the WEC).
If you're curious, here's a paper about testing a much larger version of our WEC and doing SID: https://doi.org/10.3390/en10040472
Does this clear it up?
nickross4444
The Wavemaker arduino operates most smoothly with 8 components in the jonswap, anything above begins to make the movement "jittery". I thought it may be due to computing overhead, but some other testing indicates that may not be the case. Before I spend more time on it, @gbacelli @ryancoe what is our desired number of components for the wavemaker jonswap. 8 currently is very smooth, 16 is okay but not great. I can think of a few ways to improve it, but would like to know what our target is in order to evaluate them.
Secondly, @ryancoe could you remind me(and for documentation's sake) why we are using a jonswap on the WEC controller? The current modes are direct torque control, PD controller, and Sea state, which is just an open loop jonswap timeseries output.