Zitat von: Ralf9 am 13 Mai 2026, 10:06:19Für das Senden muss ich noch was im 00_SIGNALduino... Modul ergänzen
Radio A is not active! kommt? Und kann es sein, dass ein einfaches set sduino raw SRA;;... mir derzeit meine cconfig zerschießt?Zitat von: Ralf9 am 07 Mai 2026, 20:10:55Hier ist der rfmode, eingelesen wird er mit "get raw"CW000D,022D,0307,04D3,0591,063D,0704,0832,0D21,0E6B,0FF6,1057,1143,1200,1323,14B9,1531,1700,1818,1914,1B07,1C00,1D90,23E9,242A,2500,2611,3D00,3E00,4045,4162,4249,436E,4473,4574,4661,4774
Zitat von: Ralf9 am 13 Mai 2026, 10:06:19Das müsste vorläufig auch ohne ein extra Modul gehen.
Dafür in einem dummy readings erzeugen
z.B.
123_on
123_off
456_on
456_off
Oder pro device ein dummy
Der Code dafür kann evtl da "99_myUtils.pm" abgelegt werden.
define ntf_update_dmsg notify MySignalPicoLAN:LASTDMSG:.* set myZimmer $EVTPART1
# CFGFN
# DEF MySignalPicoLAN:LASTDMSG:.* set myZimmer $EVTPART1
# FUUID 6a05829a-f33f-3e5d-77b9-f56c1e566698069a
# NAME ntf_update_dmsg
# NOTIFYDEV MySignalPicoLAN
# NR 75
# NTFY_ORDER 50-ntf_update_dmsg
# REGEXP MySignalPicoLAN:LASTDMSG:.*
# STATE active
# TYPE notify
# eventCount 1
# READINGS:
# 2026-05-14 10:06:50 state active
#
setstate ntf_update_dmsg active
setstate ntf_update_dmsg 2026-05-14 10:06:50 state active
[homebridge-fhem] This plugin generated a warning from the characteristic 'On':
characteristic value expected boolean and received undefined.bash
sudo hb-service stop
cd /var/lib/homebridge
sudo npm install --unsafe-perm homebridge@1.8.4
sudo hb-service start
bash
cat /var/lib/homebridge/node_modules/homebridge/package.json | grep '"version"'
36_Shelly.pm:v6.5.15-s31139/2026-04-21 für meinen Shelly Pro 3EMPurchased_Energy_S und Purchased_Energy_TPurchased_Energy_T habe ich einen Wert den ich absolut nicht zuordnen kannPurchased_Energy_T 1.66475359739714e+290Total_Energy_S -974800
Total_Energy_T 3.2938266762747e+289 sein, doch auch dort Werte mit denen ich nichts anfangen kann.ZitataiConAbsOversample=0.25Ja, sehr wahrscheinlich, siehe:
->> das ist wohl zu hoch von mir gewählt.
026.05.09 20:12:50 1: Forecast DEBUG> AI FANN - Absence oversampling: original=286 absent, target_ratio=25%, needed=3017, added=2731. New total=12069 records (absent share=25.0%)
2026.05.14 09:19:21 1: Forecast DEBUG> AI FANN - Absence oversampling: original=294 absent, target_ratio=25%, needed=3051, added=2757. New total=12204 records (absent share=25.0%)
2026.05.14 09:19:21 1: Forecast DEBUG> First attempt 0 with Seed=638160
2026.05.14 09:19:21 1: Forecast DEBUG> AI FANN Training started with Params:
input datasets=12204,
Registry version=v1_heatpump_active_pv,
training algo=FANN_TRAIN_INCREMENTAL,
output AF=LINEAR,
hidden AF=ELLIOT_SYMMETRIC,
hidden Neurons=80-40,
hidden steepness=1.0,
max. Epoches=15000,
mse_error=0.001,
learning rate=0.00200,
learning momentum=0.8,
BitFail limit: 0.28,
Data sharing=split after shuffle of training data and use AI internal shuffle (Train=9763, Test=2440),
Data shuffle=2 (period=20)
2026.05.14 09:19:22 1: Forecast DEBUG> Epoche 1: Train MSE=0.009878, Val MSE=0.008495, Val MAE=0.070334, Val MedAE=0.057356, Bit_Fail=24 -> Snap metric improved
2026.05.14 09:19:23 1: Forecast DEBUG> Epoche 2: Train MSE=0.006709, Val MSE=0.006639, Val MAE=0.066407, Val MedAE=0.054215, Bit_Fail=0 -> Snap metric improved
2026.05.14 09:19:24 1: Forecast DEBUG> Epoche 3: Train MSE=0.004458, Val MSE=0.005276, Val MAE=0.059123, Val MedAE=0.053290, Bit_Fail=0 -> Snap metric improved
2026.05.14 09:19:25 1: Forecast DEBUG> Epoche 4: Train MSE=0.003945, Val MSE=0.004404, Val MAE=0.053285, Val MedAE=0.045342, Bit_Fail=0 -> Snap metric improved
2026.05.14 09:34:15 1: Forecast DEBUG> AI FANN - Absence oversampling: original=294 absent, target_ratio=10%, needed=1017, added=723. New total=10170 records (absent share=10.0%)
2026.05.14 09:34:16 1: Forecast DEBUG> First attempt 0 with Seed=956776
2026.05.14 09:34:16 1: Forecast DEBUG> AI FANN Training started with Params:
input datasets=10170,
Registry version=v1_heatpump_active_pv,
training algo=FANN_TRAIN_INCREMENTAL,
output AF=LINEAR,
hidden AF=ELLIOT_SYMMETRIC,
hidden Neurons=80-40,
hidden steepness=1.0,
max. Epoches=15000,
mse_error=0.001,
learning rate=0.00200,
learning momentum=0.8,
BitFail limit: 0.28,
Data sharing=split after shuffle of training data and use AI internal shuffle (Train=8136, Test=2033),
Data shuffle=2 (period=20)
2026.05.14 09:34:17 1: Forecast DEBUG> Epoche 1: Train MSE=0.010520, Val MSE=0.010959, Val MAE=0.087255, Val MedAE=0.083240, Bit_Fail=18 -> Snap metric improved
2026.05.14 09:34:17 1: Forecast DEBUG> Epoche 2: Train MSE=0.007831, Val MSE=0.009136, Val MAE=0.077880, Val MedAE=0.070504, Bit_Fail=5 -> Snap metric improved
2026.05.14 09:34:18 1: Forecast DEBUG> Epoche 3: Train MSE=0.005414, Val MSE=0.007079, Val MAE=0.067047, Val MedAE=0.058759, Bit_Fail=3 -> Snap metric improved
2026.05.14 09:34:19 1: Forecast DEBUG> Epoche 4: Train MSE=0.004304, Val MSE=0.005714, Val MAE=0.060038, Val MedAE=0.051889, Bit_Fail=2 -> Snap metric improved
2026.05.14 09:34:19 1: Forecast DEBUG> Epoche 5: Train MSE=0.004044, Val MSE=0.005043, Val MAE=0.056088, Val MedAE=0.047296, Bit_Fail=2 -> Snap metric improved
2026.05.14 09:34:20 1: Forecast DEBUG> Epoche 6: Train MSE=0.003946, Val MSE=0.004730, Val MAE=0.053946, Val MedAE=0.044954, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:21 1: Forecast DEBUG> Epoche 7: Train MSE=0.003890, Val MSE=0.004557, Val MAE=0.052602, Val MedAE=0.043257, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:21 1: Forecast DEBUG> Epoche 8: Train MSE=0.003849, Val MSE=0.004444, Val MAE=0.051639, Val MedAE=0.042007, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:22 1: Forecast DEBUG> Epoche 9: Train MSE=0.003816, Val MSE=0.004363, Val MAE=0.050882, Val MedAE=0.041097, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:22 1: Forecast DEBUG> Epoche 10: Train MSE=0.003789, Val MSE=0.004301, Val MAE=0.050253, Val MedAE=0.039772, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:23 1: Forecast DEBUG> Epoche 11: Train MSE=0.003766, Val MSE=0.004250, Val MAE=0.049714, Val MedAE=0.038807, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:24 1: Forecast DEBUG> Epoche 12: Train MSE=0.003745, Val MSE=0.004208, Val MAE=0.049244, Val MedAE=0.038120, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:24 1: Forecast DEBUG> Epoche 13: Train MSE=0.003727, Val MSE=0.004174, Val MAE=0.048835, Val MedAE=0.037742, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:25 1: Forecast DEBUG> Epoche 14: Train MSE=0.003710, Val MSE=0.004146, Val MAE=0.048476, Val MedAE=0.037400, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:26 1: Forecast DEBUG> Epoche 15: Train MSE=0.003694, Val MSE=0.004122, Val MAE=0.048161, Val MedAE=0.036715, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:26 1: Forecast DEBUG> Epoche 16: Train MSE=0.003680, Val MSE=0.004103, Val MAE=0.047885, Val MedAE=0.036138, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:27 1: Forecast DEBUG> Epoche 17: Train MSE=0.003666, Val MSE=0.004087, Val MAE=0.047641, Val MedAE=0.035685, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:28 1: Forecast DEBUG> Epoche 18: Train MSE=0.003653, Val MSE=0.004074, Val MAE=0.047428, Val MedAE=0.035457, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:28 1: Forecast DEBUG> Epoche 19: Train MSE=0.003641, Val MSE=0.004064, Val MAE=0.047242, Val MedAE=0.035230, Bit_Fail=1 -> Snap metric improved
2026.05.14 09:34:30 1: Forecast DEBUG> Epoche 21: Train MSE=0.005799, Val MSE=0.004635, Val MAE=0.048877, Val MedAE=0.034547, Bit_Fail=1 -> Snap weighted rmse improved
2026.05.14 09:34:30 1: Forecast DEBUG> Epoche 22: Train MSE=0.005641, Val MSE=0.004596, Val MAE=0.048539, Val MedAE=0.034717, Bit_Fail=1 -> Snap weighted rmse improved
2026.05.14 09:34:31 1: Forecast DEBUG> Epoche 23: Train MSE=0.005519, Val MSE=0.004600, Val MAE=0.048449, Val MedAE=0.033969, Bit_Fail=1 -> Snap weighted rmse improved
2026.05.14 09:34:32 1: Forecast DEBUG> Epoche 24: Train MSE=0.005410, Val MSE=0.004629, Val MAE=0.048512, Val MedAE=0.033811, Bit_Fail=1 -> Snap weighted rmse improved
2026.05.14 09:34:33 1: Forecast DEBUG> Epoche 26: Train MSE=0.005211, Val MSE=0.004744, Val MAE=0.048981, Val MedAE=0.033613, Bit_Fail=1 -> Snap weighted rmse improved
2026.05.14 09:35:22 1: Forecast DEBUG> Epoche 100: Train MSE=0.003605, Val MSE=0.009403, Val MAE=0.071706, Val MedAE=0.049997, Bit_Fail=32
2026.05.14 09:36:29 1: Forecast DEBUG> Epoche 200: Train MSE=0.003056, Val MSE=0.008448, Val MAE=0.063368, Val MedAE=0.039779, Bit_Fail=27
2026.05.14 09:37:01 1: PERL WARNING: Argument "" isn't numeric in sprintf at (eval 259912) line 1.
2026.05.14 09:37:35 1: Forecast DEBUG> Epoche 300: Train MSE=0.002764, Val MSE=0.010413, Val MAE=0.075234, Val MedAE=0.054856, Bit_Fail=34
2026.05.14 09:38:39 1: Forecast DEBUG> Epoche 400: Train MSE=0.002583, Val MSE=0.007742, Val MAE=0.060097, Val MedAE=0.037552, Bit_Fail=25
2026.05.14 09:39:44 1: Forecast DEBUG> Epoche 500: Train MSE=0.002410, Val MSE=0.008633, Val MAE=0.064715, Val MedAE=0.042794, Bit_Fail=35
2026.05.14 09:40:49 1: Forecast DEBUG> Epoche 600: Train MSE=0.002265, Val MSE=0.009091, Val MAE=0.064608, Val MedAE=0.040543, Bit_Fail=40
2026.05.14 09:41:54 1: Forecast DEBUG> Epoche 700: Train MSE=0.002152, Val MSE=0.008678, Val MAE=0.063506, Val MedAE=0.039698, Bit_Fail=38
2026.05.14 09:42:01 1: PERL WARNING: Argument "" isn't numeric in sprintf at (eval 274625) line 1.
2026.05.14 09:42:59 1: Forecast DEBUG> Epoche 800: Train MSE=0.002042, Val MSE=0.009026, Val MAE=0.064545, Val MedAE=0.040666, Bit_Fail=46
2026.05.14 09:44:04 1: Forecast DEBUG> Epoche 900: Train MSE=0.001937, Val MSE=0.008428, Val MAE=0.062749, Val MedAE=0.040911, Bit_Fail=39
2026.05.14 09:45:10 1: Forecast DEBUG> Epoche 1000: Train MSE=0.001850, Val MSE=0.008596, Val MAE=0.063272, Val MedAE=0.040617, Bit_Fail=42
2026.05.14 09:45:27 1: Forecast DEBUG> Early stopping bei Epoche 1026 (no improvement since 1000 epochs)
2026.05.14 09:45:27 1: === Snapshot-Statistik ===
2026.05.14 09:45:27 1: Metric-Improvement Snapshots: 19 (letzte Epoche: 19)
2026.05.14 09:45:27 1: Weighted-RMSE-Proxy-Improvement Snapshots: 5 (letzte Epoche: 26)
2026.05.14 09:45:27 1: Bit-Improvement Snapshots: 0 (letzte Epoche: 0)
2026.05.14 09:45:27 1: Bit-Tradeoff Snapshots: 0 (letzte Epoche: 0)
2026.05.14 09:45:27 1: Forecast DEBUG> Best Snapshot reloaded from Epoche 26: Train MSE=0.005211, Val MSE=0.004744, Val MAE=0.048981, Val MedAE=0.033613, Bit_Fail=1,
2026.05.14 09:45:27 1: Forecast DEBUG> Run Validation Test with 20% of Input data ...
2026.05.14 09:45:27 1: Forecast DEBUG> Validation finished - Best Training MSE=0.005211, Validation MSE=0.004744, Validation Bit_Fail=1
2026.05.14 09:45:27 1: Forecast DEBUG> Retrain check ->
Zitatdie grauen Balken sind consumptionForecast und die blauen gridconsumption. Jedoch wird bei gridconsumption der Eigenverbrauch von der PV mitgerechnet, ebenso der Teil der in den Akku gehtgridconsumption ist ein Wert der direkt aus dem gemessenen bzw. gelieferten Wert von setupMeterDev->contotal abgeleitet wird. Er ist ebenfalls in den Readings Today_HourXX_GridConsumption ersichtlich.
2026.05.14 09:47:49.718 1: SolCast DEBUG> collect Energy Meter data - device: SMA_Energymeter =>
2026.05.14 09:47:49.718 1: SolCast DEBUG> gcon: 19.7 W, gfeedin: 0 W, contotal: 3497.6 Wh, feedtotal: 37117.3 Wh
Relevant ist gcon für den aktuellen Bezug, dargestellt in der Flußgrafik. Der Wert in contotal ist der totale Energiebezug aus dem der Stundenwert abgeleitet wird. Dieser Wert wäre in deinem beschriebenen Kontext relevant.<ftui-chart-data fill type="bar" background-color="rgba(255, 186, 8 ,0.5)" color="rgba(250, 163, 7,1)" stack="stack1" log="lgprxy" spec="Func:myPVplot2(2)" [update]="PVStats:button_pvforecasttmrrw" [hidden]="PVStats:button_pvforecasttmrrw | map('off:false,on:true')"></ftui-chart-data>
hidden="true" scheint zu funktionieren, aber ein hidden="false" sorgt trotzdem dafür, dass der entsprechende Datensatz ausgeblendet wird.[style]="PVStats:button_pvforecasttmrrw | map('on:``, off:`display: none;`')"
funktioniert auch nicht (mehr?).