DPL: Solar inverter max increase calculcation is less strict, respects overscaling configuration and more

[AndreasBoehm]

As we are having issues with solar inverters reporting that they can't increase their power i tried to analyse the methods and developed ideas what could be the cause and how to fix those. + the configured overscaling input power threshold should be used for this calculation as well.

Ideas

Low limits

Inverters with low limit (below 15%) are usually not able to provide the expected power and we would rate mppts as shaded.
This will result in us thinking that we can't get more power from that inverter and the limit is 'stuck'.

Strict checks

The max increase calculation is based on the overscaling method but it actually does not need to be as strict.
I lowered the input power threshold from 98% to 97% to be less strict and to hopefully let more inverters recover from low limits.
This won't hurt because if the inverter is actually not able to provide what we calculated we will fix this in the next DPL run.

Threshold config

As said earlier the max increase calculation is based on the overscaling method, and it needs to be, we need to make sure they stay aligned. Thats why i will now use the configured input threshold to calculate max increase.

Missing visibility

We can't really debug whats going on because the key factor to determine the max increase and also for scaling the limit is the power power MPPT. I added this information when debug logging is enabled.

11:15:01.014 > [DPL inverter XXXXXXXXX]:
11:15:01.016 >     battery-powered, producing 195 W
11:15:01.019 >     lower/current/upper limit: 25/200/2000 W, output capability: 2000 W
11:15:01.022 >     sending commands enabled, reachable, eligible
11:15:01.028 >     max reduction production/standby: 170/195 W, max increase: 1800 W
11:15:01.030 >     target limit/output/state: 205 W (update)/205 W/production, 0 update timeouts
11:15:01.033 >     MPPTs AC power: a: 49 W b: 49 W c: 49 W d: 48 W

Related discussion

• DPL (Dynamic Power LImit) Probleme mit Regelung eines HMS-1600-4T #1467

[schlimmchen]

Looks good to me. The commit summary of that one (oldest or newest, not sure) commit is extremely long and it will appear as such in the changelogs... Not ideal. Maybe next time you can try harder to find a shorter description or split the commit.

[AndreasBoehm]

Sorry, I will try to make shorter commit summaries in the future.