2026.01.0

dao/CHANGELOG.md

@@ -1,6 +1,17 @@
 # Changelog 刀 DAO
 # Day Ahead Optimizer
-# 2025.12.1
+ # 2026.01.0
+With this new stable release comes a second optional method for predicting the porduction of your solar-devices.
+This if fully documented in the wiki: https://github.com/corneel27/day-ahead/wiki/2.-Installatie-en-basis-configuratie#pv-productie-voorspellen <br>
+This module is developed by @simnet and implemented by @corneel27/kc27.
+Other changes and fixes:
+- Updated several used python packages
+- charge and discharge of the battery is now fully calculated with "specializes order set" (sos). 
+This gives always the "optimal" (=most efficient) level voor charging and discharging.
+- Update README.md, link to wiki: https://github.com/corneel27/day-ahead/wiki (thanks @Torch1969)
+- From now on new features will only be documented in the wiki, in the future DOCS.md will become deprecated.
+
+ # 2025.12.1
 - Fix error calculating number of heat-blocks heatpump (on/off, reported by @sailor_dg)
 
 # 2025.12.0

dao/config.yaml

@@ -1,6 +1,6 @@
 ---
 name: 刀 Day Ahead Optimizer
-version: 2025.12.1
+version: 2026.01.0
 slug: day_ahead_opt
 description: Home Assistant Community Add-ons for day ahead optimizations
 url: https://github.com/corneel27/day-ahead

dao/prog/da_meteo.py

@@ -182,7 +182,7 @@ def get_dif_rad_factor(self, utc_time):
         return value
 
     @staticmethod
-    def is_aws(station:int):
+    def is_aws(station: int):
         """
         station :code van een knmi station
         :return: boolean
@@ -199,8 +199,39 @@ def is_aws(station:int):
         """
         # onderstaande lijst is gegenereerd met prof/tst.py/generate_list_knmi-aws.py
         # beter bij iedere nieuwe versie autoamtisch checken en vernieuwen op github
-        list_aws = [215, 235, 240, 249, 251, 257, 260, 267, 269, 270, 273, 275, 277, 278, 279, 280,
-                    283, 286, 290, 310, 319, 323, 330, 344, 348, 350, 356, 370, 375, 377, 380]
+        list_aws = [
+            215,
+            235,
+            240,
+            249,
+            251,
+            257,
+            260,
+            267,
+            269,
+            270,
+            273,
+            275,
+            277,
+            278,
+            279,
+            280,
+            283,
+            286,
+            290,
+            310,
+            319,
+            323,
+            330,
+            344,
+            348,
+            350,
+            356,
+            370,
+            375,
+            377,
+            380,
+        ]
         return station in list_aws
 
     def which_station(self) -> str:

dao/prog/da_report.py

@@ -1145,15 +1145,19 @@ def add_col_df(
                     except ValueError:
                         org_value = 0
                     if pd.notna(row[col_index]):
-                        add_to.at[row.tijd, col_name_to] = org_value + factor * row[col_index]
+                        add_to.at[row.tijd, col_name_to] = (
+                            org_value + factor * row[col_index]
+                        )
         else:
             for row in add_from.itertuples():
                 # add_from.at[row.tijd, col_name_from])
                 if row.time in add_to.index:
                     org_value = add_to.at[row.tijd, col_name_to]
                     if pd.isna(org_value):
                         org_value = 0
-                    add_to.at[row.time, col_name_to] = org_value + factor * row[col_index]
+                    add_to.at[row.time, col_name_to] = (
+                        org_value + factor * row[col_index]
+                    )
         return add_to
 
     def get_latest_present(self, code: str) -> datetime.datetime:
@@ -2998,7 +3002,7 @@ def calc_solar_data(self, device: dict, day: datetime.date, active_view: str):
 
         # prognose straling
         rad_prog = self.get_da_data("gr", start, end, "uur", "uur", "prognoses")
-        result["prognose_straling"] =rad_prog["gr"]
+        result["prognose_straling"] = rad_prog["gr"]
 
         # gemeten straling
         rad_real = self.get_da_data("gr", start, end, "uur", "uur", "values")
@@ -3017,7 +3021,7 @@ def calc_solar_data(self, device: dict, day: datetime.date, active_view: str):
                 self.add_col_df(df_sensor, df_solar, "gemeten")
             count += 1
         result["gemeten_prod"] = pd.NA
-        self.add_col_df(df_solar, result, "gemeten", "gemeten_prod" )
+        self.add_col_df(df_solar, result, "gemeten", "gemeten_prod")
 
         # voorspelling DAO
         pred_dao = []
@@ -3039,12 +3043,13 @@ def calc_solar_data(self, device: dict, day: datetime.date, active_view: str):
 
         # voorspelling ML
         from dao.prog.solar_predictor import SolarPredictor
+
         solar_predictor = SolarPredictor()
         solar_prog = solar_predictor.predict_solar_device(device, start, end)
         solar_prog["tijd"] = solar_prog["date_time"].dt.tz_localize(None)
         solar_prog.index = solar_prog["tijd"]
         result["prognose_ml"] = pd.NA
-        self.add_col_df(solar_prog, result, "prediction","prognose_ml")
+        self.add_col_df(solar_prog, result, "prediction", "prognose_ml")
         # result["prognose_ml"] = solar_prog["prediction"]
 
         result.drop(columns=["tijd"], inplace=True)

dao/prog/day_ahead.py

@@ -709,12 +709,7 @@ def calc_optimum(
 
         # vanaf hier declaraties ontladen met sos  ###############################################
         ac_from_dc_samples = [
-            [
-                (
-                    discharge_stages[b][ds]["power"]
-                    / 1000
-                )
-                for ds in range(DS[b])]
+            [(discharge_stages[b][ds]["power"] / 1000) for ds in range(DS[b])]
             for b in range(B)
         ]
         dc_to_ac_samples = [
@@ -3192,21 +3187,29 @@ def calc_optimum(
                 if self.log_level == logging.INFO:
                     # debug laden
                     if ac_to_dc[b][u].x > 0.0:
-                        logging.info(f"Laad volume in uur {u} {uur[u]} "
-                                     f"{ac_from_dc[b][u].x*hour_fraction[u]} kWh")
+                        logging.info(
+                            f"Laad volume in uur {u} {uur[u]} "
+                            f"{ac_from_dc[b][u].x * hour_fraction[u]} kWh"
+                        )
                         for cs in range(CS[b]):
                             if ac_to_dc_w[b][u][cs].x > 0:
-                                logging.info(f"{cs} {ac_to_dc_w[b][u][cs].x} "
-                                             f"{ac_to_dc_samples[b][cs]}")
+                                logging.info(
+                                    f"{cs} {ac_to_dc_w[b][u][cs].x} "
+                                    f"{ac_to_dc_samples[b][cs]}"
+                                )
 
                     # debug ontladen
                     if ac_from_dc[b][u].x > 0.0:
-                        logging.info(f"Ontlaad volume in uur {u} {uur[u]} "
-                                     f"{ac_from_dc[b][u].x*hour_fraction[u]} kWh")
+                        logging.info(
+                            f"Ontlaad volume in uur {u} {uur[u]} "
+                            f"{ac_from_dc[b][u].x * hour_fraction[u]} kWh"
+                        )
                         for ds in range(DS[b]):
                             if ac_from_dc_w[b][u][ds].x > 0:
-                                logging.info(f"{ds} {ac_from_dc_w[b][u][ds].x} "
-                                             f"{ac_from_dc_samples[b][ds]}")
+                                logging.info(
+                                    f"{ds} {ac_from_dc_w[b][u][ds].x} "
+                                    f"{ac_from_dc_samples[b][ds]}"
+                                )
 
                 row = [
                     str(uur[u]),
@@ -3727,7 +3730,7 @@ def calc_optimum(
                         netto_vermogen_bat = -minimum_power
                 elif ac_to_dc[b][0].x > 0.0:  # laden met optimaal vermogen
                     sum_weight_factor = 0
-                    sum_power = 0 # in W
+                    sum_power = 0  # in W
                     for cs in range(CS[b]):
                         wf = ac_to_dc_w[b][0][cs].x
                         if wf > 0:
@@ -3736,36 +3739,37 @@ def calc_optimum(
                     if sum_weight_factor < 0.95:
                         new_state = battery_state_on_value
                         balance = False
-                        netto_vermogen_bat = round(sum_power/sum_weight_factor)
+                        netto_vermogen_bat = round(sum_power / sum_weight_factor)
                         new_ts = (
-                            start_dt.timestamp() + self.interval_s * sum_weight_factor * interval_fraction[0]
+                            start_dt.timestamp()
+                            + self.interval_s * sum_weight_factor * interval_fraction[0]
                         )
                         stop_omvormer = dt.datetime.fromtimestamp(int(new_ts))
                     else:
                         new_state = battery_state_on_value
                         balance = False
                         stop_omvormer = None
                 elif ac_from_dc[b][0].x > 0.0:  # ontladen met optimaal vermogen
-                        sum_weight_factor = 0
-                        sum_power = 0  # in W
-                        for ds in range(DS[b]):
-                            wf = ac_from_dc_w[b][0][ds].x
-                            if wf > 0:
-                                sum_weight_factor += wf
-                                sum_power += wf * discharge_stages[b][ds]["power"]
-                        if sum_weight_factor < 0.95:
-                            new_state = battery_state_on_value
-                            balance = False
-                            netto_vermogen_bat = - round(sum_power / sum_weight_factor)
-                            new_ts = (
-                                    start_dt.timestamp() + self.interval_s * sum_weight_factor *
-                                    interval_fraction[0]
-                            )
-                            stop_omvormer = dt.datetime.fromtimestamp(int(new_ts))
-                        else:
-                            new_state = battery_state_on_value
-                            balance = False
-                            stop_omvormer = None
+                    sum_weight_factor = 0
+                    sum_power = 0  # in W
+                    for ds in range(DS[b]):
+                        wf = ac_from_dc_w[b][0][ds].x
+                        if wf > 0:
+                            sum_weight_factor += wf
+                            sum_power += wf * discharge_stages[b][ds]["power"]
+                    if sum_weight_factor < 0.95:
+                        new_state = battery_state_on_value
+                        balance = False
+                        netto_vermogen_bat = -round(sum_power / sum_weight_factor)
+                        new_ts = (
+                            start_dt.timestamp()
+                            + self.interval_s * sum_weight_factor * interval_fraction[0]
+                        )
+                        stop_omvormer = dt.datetime.fromtimestamp(int(new_ts))
+                    else:
+                        new_state = battery_state_on_value
+                        balance = False
+                        stop_omvormer = None
                 else:
                     new_state = battery_state_on_value
                     balance = False

dao/prog/solar_predictor.py

@@ -1037,7 +1037,7 @@ def predict_solar_device(
                 f"Er is geen model aanwezig voor {self.solar_name},svp eerst trainen."
             )
         latest_dt = self.db_da.get_time_border_record("gr", latest=True)
-        prognose = (latest_dt < end)
+        prognose = latest_dt < end
         weather_data = self.get_weatherdata(start, end, prognose=prognose)
         prediction = self.predict(weather_data)
         logging.info(f"ML prediction {self.solar_name}\n{prediction}")

release-testing/CHANGELOG.md

@@ -1,5 +1,8 @@
 # Changelog 刀 DAO
 # Day Ahead Optimizer
+# 2026.01.0.rc9
+This version is identical to stable version 2026.01.0
+
 # 2026.01.0.rc8
 Fixed error check AWS (reported by @Kees Schilder)
 
@@ -135,9 +138,9 @@ When you activate this rule: for devices with ml_prediction = True there will be
 9. In the dashboard is an extra-menu-option available: **Solar**
 Here you can view the effectivity of your calculated model:<br>
 In text:<br>
-![solar_table.png](images/solar_table.png)<br>
+![solar_table.png](https://github.com/corneel27/day-ahead/raw/main/dao/images/solar_table.png)<br>
 With a graph:<br>
-![solar_graph.png](/images/solar_graph.png)
+![solar_graph.png](https://github.com/corneel27/day-ahead/raw/main/dao//images/solar_graph.png)
 
 # 2025.12.1.rc1
 - Fix error calculating number of heat-blocks heatpump (on/off, reported by @sailor_dg)

release-testing/config.yaml

@@ -1,6 +1,6 @@
 ---
 name: 刀 Day Ahead Optimizer (TESTING)
-version: 2026.01.0.rc8
+version: 2026.01.0.rc9
 stage: experimental
 slug: day_ahead_opt-testing
 description: Beta version of DAO. Use only for testing!