Enabling the new Startup Penalty function

[purcell-lab]

@werdnum thanks for your contribution with the startup penalty function.

I too am running a pool filter pump & heat pump against Amber pricing so this feature could be useful, I currently have a delay automation so I don't call things more than every 15 minutes. With your start penalty function if I really need to stop or start without waiting the 15 minutes I still can.

I have dropped in some default values, but am interesting in tuning them:

        "def_current_state": {{[
                        iif(states('sensor.p_deferrable0')|int(0),1,0),
                        iif(states('sensor.p_deferrable1')|int(0),1,0),
                        iif(states('sensor.p_deferrable2')|int(0),1,0),
                        iif(states('sensor.p_deferrable3')|int(0),1,0),
                        iif(states('sensor.p_deferrable4')|int(0),1,0),
                        iif(states('sensor.p_deferrable5')|int(0),1,0)
                      ]}},
        "def_start_penalty":      [1, 1, 1, 1, 1, 1],

I couldn't see any debug entries (like the optimisation windows).

How can I tell the penalties have been applied and where would I see an effect?

2024-06-12 23:00:24,397 - web_server - INFO - Passed runtime parameters: {'load_cost_forecast': [0.29, 0.25, 0.25, 0.22, 0.24, 0.2, 0.18, 0.19, 0.17, 0.19, 0.17, 0.17, 0.17, 0.21, 0.44, 0.51, 0.9, 0.52, 0.41, 0.15, 0.12, 0.11, 0.05, 0.05, 0.05, 0.04, 0.04, 0.04, 0.04, 0.05, 0.06, 0.06, 0.12, 0.12, 0.29, 0.64, 8.06, 17.91, 17.91, 17.91, 16.32, 1.73, 1.23, 0.98, 0.52, 0.35, 0.44, 0.47, 0.44, 0.4, 0.34, 0.19, 0.21, 0.2, 0.2, 0.2, 0.18, 0.17], 'prod_price_forecast': [0.19, 0.16, 0.16, 0.13, 0.15, 0.1, 0.09, 0.1, 0.08, 0.1, 0.08, 0.08, 0.08, 0.12, 0.33, 0.39, 0.74, 0.4, 0.3, 0.06, 0.04, 0.04, -0.02, -0.02, -0.02, -0.03, -0.03, -0.03, -0.03, -0.02, -0.01, -0.01, 0.04, 0.04, 0.07, 0.39, 7.13, 16.09, 16.09, 16.09, 14.63, 1.37, 0.92, 0.7, 0.4, 0.25, 0.33, 0.35, 0.33, 0.29, 0.24, 0.1, 0.11, 0.11, 0.11, 0.11, 0.09, 0.08], 'pv_power_forecast': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 60, 286, 1355, 3001, 4483, 5779, 6988, 7995, 8730, 9235, 9494, 9574, 9375, 8958, 8300, 7425, 6303, 4985, 3562, 1964, 539, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 107, 422, 1397, 2782, 4093, 5358, 6664, 7754, 8566, 9058, 9301, 9375, 9196, 8799, 8154, 7165, 5796, 4525, 3149, 1643, 348, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 'load_power_forecast': [1120, 1100, 400, 1100, 1100, 1000, 1300, 1400, 1300, 2000, 1500, 1600, 2000, 1600, 1700, 6800, 1500, 1600, 2000, 1700, 1800, 300, 300, 600, 1200, 1600, 2800, 1600, 400, 300, 2200, 500, 1000, 300, 300, 500, 400, 400, 600, 700, 1300, 1400, 1200, 1100, 1000, 1000, 1100, 1100, 1100, 1100, 400, 1100, 1100, 1000, 1300, 1400, 1300, 2000, 1500, 1600, 2000, 1600, 1700, 6800, 1500, 1600, 2000, 1700, 1800, 300, 300, 600, 1200, 1600, 2800, 1600, 400, 300, 2200, 500, 1000, 300, 300, 500, 400, 400, 600, 700, 1300, 1400, 1200, 1100, 1000, 1000, 1100, 1100], 'prediction_horizon': 58, 'alpha': 1, 'beta': 0, 'num_def_loads': 6, 'def_total_hours': [0, 5, 0, 0, 0, 0], 'def_end_timestep': [0, 0, 0, 0, 15, 0], 'P_deferrable_nom': [1232, 6333, 11520, 2500, 600, 11520], 'def_current_state': [0, 0, 0, 0, 0, 0], 'def_start_penalty': [1, 1, 1, 1, 1, 1], 'treat_def_as_semi_cont': [1, 1, 0, 0, 1, 0], 'set_def_constant': [0, 0, 0, 0, 0, 0], 'soc_init': 0.14, 'soc_final': 0}
2024-06-12 23:00:24,398 - web_server - INFO -  >> Setting input data dict
2024-06-12 23:00:24,398 - web_server - INFO - Setting up needed data
2024-06-12 23:00:24,404 - web_server - INFO - Retrieve hass get data method initiated...
2024-06-12 23:00:24,940 - web_server - INFO - Retrieving weather forecast data using method = list
2024-06-12 23:00:24,949 - web_server - INFO -  >> Performing naive MPC optimization...
2024-06-12 23:00:24,949 - web_server - INFO - Performing naive MPC optimization
2024-06-12 23:00:24,987 - web_server - INFO - Perform an iteration of a naive MPC controller
2024-06-12 23:00:25,199 - web_server - DEBUG - Deferrable load 0: Proposed optimization window: 0 --> 0
2024-06-12 23:00:25,199 - web_server - DEBUG - Deferrable load 0: Validated optimization window: 0 --> 0
2024-06-12 23:00:25,216 - web_server - DEBUG - Deferrable load 1: Proposed optimization window: 0 --> 0
2024-06-12 23:00:25,216 - web_server - DEBUG - Deferrable load 1: Validated optimization window: 0 --> 0
2024-06-12 23:00:25,233 - web_server - DEBUG - Deferrable load 2: Proposed optimization window: 0 --> 0
2024-06-12 23:00:25,233 - web_server - DEBUG - Deferrable load 2: Validated optimization window: 0 --> 0
2024-06-12 23:00:25,241 - web_server - DEBUG - Deferrable load 3: Proposed optimization window: 0 --> 0
2024-06-12 23:00:25,241 - web_server - DEBUG - Deferrable load 3: Validated optimization window: 0 --> 0
2024-06-12 23:00:25,254 - web_server - DEBUG - Deferrable load 4: Proposed optimization window: 0 --> 15
2024-06-12 23:00:25,254 - web_server - DEBUG - Deferrable load 4: Validated optimization window: 0 --> 15
2024-06-12 23:00:25,272 - web_server - DEBUG - Deferrable load 5: Proposed optimization window: 0 --> 0
2024-06-12 23:00:25,273 - web_server - DEBUG - Deferrable load 5: Validated optimization window: 0 --> 0
2024-06-12 23:00:26,354 - web_server - INFO - Status: Optimal
2024-06-12 23:00:26,354 - web_server - INFO - Total value of the Cost function = 476.25
2024-06-12 23:00:27,174 - web_server - INFO - Passed runtime parameters: {'custom_load_forecast_id': {'entity_id': 'sensor.p_load_forecast', 'unit_of_measurement': 'W', 'friendly_name': 'Load Power Forecast'}}
2024-06-12 23:00:27,174 - web_server - INFO -  >> Setting input data dict
2024-06-12 23:00:27,174 - web_server - INFO - Setting up needed data
2024-06-12 23:00:27,178 - web_server - INFO -  >> Publishing data...
2024-06-12 23:00:27,179 - web_server - INFO - Publishing data to HASS instance
2024-06-12 23:00:27,215 - web_server - INFO - Successfully posted to sensor.p_pv_forecast = 0
2024-06-12 23:00:27,243 - web_server - INFO - Successfully posted to sensor.p_load_forecast = 1120
2024-06-12 23:00:27,268 - web_server - INFO - Successfully posted to sensor.p_pv_curtailment = 0.0
2024-06-12 23:00:27,295 - web_server - INFO - Successfully posted to sensor.p_deferrable0 = 0.0
2024-06-12 23:00:27,322 - web_server - INFO - Successfully posted to sensor.p_deferrable1 = 0.0
2024-06-12 23:00:27,350 - web_server - INFO - Successfully posted to sensor.p_deferrable2 = 0.0
2024-06-12 23:00:27,386 - web_server - INFO - Successfully posted to sensor.p_deferrable3 = 0.0
2024-06-12 23:00:27,410 - web_server - INFO - Successfully posted to sensor.p_deferrable4 = 0.0
2024-06-12 23:00:27,437 - web_server - INFO - Successfully posted to sensor.p_deferrable5 = 0.0
2024-06-12 23:00:27,471 - web_server - INFO - Successfully posted to sensor.p_batt_forecast = 0.0
2024-06-12 23:00:27,499 - web_server - INFO - Successfully posted to sensor.soc_batt_forecast = 14.0
2024-06-12 23:00:27,530 - web_server - INFO - Successfully posted to sensor.p_grid_forecast = 1120.0
2024-06-12 23:00:27,557 - web_server - INFO - Successfully posted to sensor.total_cost_fun_value = 486.26
2024-06-12 23:00:27,577 - web_server - INFO - Successfully posted to sensor.optim_status = Optimal
2024-06-12 23:00:27,607 - web_server - INFO - Successfully posted to sensor.unit_load_cost = 0.29
2024-06-12 23:00:27,635 - web_server - INFO - Successfully posted to sensor.unit_prod_price = 0.19

I'm still seeing a 30 minute gap in the scheduling of my pool heater for tomorrow, which is probably OK. I'm hopeful this startup_penalty will prevent my high frequency MPC switching on/ off each minute. But how do I know it's functional?

Pool startup's look a lot smoother today compared to yesterday. Deferrable1 yesterday was starting and stopping a lot, today there has only been one start/stop cycle:

[mime24]

Hello everyone, I have configured two thermal loads (def0 = hot water and def3 = heating) and have been trying for days to influence the frequency of starts using the penalty parameter, but without success. I have tried various numbers between 1 and 100, but several starts are often scheduled in quick succession.

Do you have any tips?

here multiple starte in the night

al lot of starts

sometimes it looks fine like here

[2025-10-10 08:55:20 +0200] [21] [INFO] Passed runtime parameters: {'method': 'mlforecaster', 'prediction_horizon': 96, 'entity_save': True, 'publish_prefix': 'emhass_', 'soc_init': 0.262, 'soc_final': 0.5, 'load_cost_forecast': [0.3446, 0.4015, 0.3637, 0.3455, 0.3237, 0.3767, 0.3412, 0.331, 0.3152, 0.3449, 0.3247, 0.3244, 0.3183, 0.3349, 0.3254, 0.3192, 0.311, 0.3209, 0.3242, 0.3223, 0.3144, 0.3017, 0.3225, 0.3256, 0.3394, 0.3214, 0.3316, 0.3393, 0.3523, 0.3221, 0.3393, 0.3492, 0.3767, 0.3467, 0.3635, 0.3772, 0.3967, 0.3579, 0.3702, 0.3748, 0.3926, 0.3789, 0.3777, 0.3778, 0.3739, 0.3873, 0.3724, 0.3613, 0.3498, 0.3678, 0.3624, 0.3506, 0.3352, 0.3548, 0.3523, 0.3469, 0.3327, 0.3429, 0.341, 0.3453, 0.3365, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35, 0.35], 'operating_hours_of_each_deferrable_load': [0, 0, 0, 0, 1], 'def_current_state': [False, False, False, False, False], 'def_start_penalty': [100, 5, 1, 100, 1], 'start_timesteps_of_each_deferrable_load': [0, 0, 0, 0, 36], 'end_timesteps_of_each_deferrable_load': [24, 26, 0, 0, 40], 'nominal_power_of_deferrable_loads': [1500, 300, 1000, 450, 1000], 'minimum_power_of_deferrable_loads': [1400, 0, 0, 440, 0], 'def_load_config': [{'thermal_config': {'heating_rate': 15, 'cooling_constant': 0.004, 'overshoot_temperature': 48.0, 'start_temperature': 39.9, 'desired_temperatures': [32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 45.0, 45.0, 45.0, 45.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0, 32.0]}}, {}, {}, {'thermal_config': {'heating_rate': 5, 'cooling_constant': 0.018, 'overshoot_temperature': 26.0, 'start_temperature': 23.4, 'desired_temperatures': [21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.0, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5, 21.5]}}, {}], 'outdoor_temperature_forecast': [8.73, 9.3, 9.88, 10.45, 11.03, 11.6, 12.0, 12.4, 12.8, 13.2, 13.5, 13.8, 14.1, 14.4, 14.62, 14.85, 15.08, 15.3, 15.33, 15.35, 15.38, 15.4, 15.4, 15.4, 15.4, 15.4, 15.45, 15.5, 15.55, 15.6, 15.55, 15.5, 15.45, 15.4, 15.12, 14.85, 14.58, 14.3, 13.65, 13.0, 12.35, 11.7, 11.62, 11.55, 11.47, 11.4, 11.2, 11.0, 10.8, 10.6, 10.5, 10.4, 10.3, 10.2, 10.1, 10.0, 9.9, 9.8, 9.68, 9.55, 9.43, 9.3, 9.15, 9.0, 8.85, 8.7, 8.67, 8.65, 8.62, 8.6, 8.62, 8.65, 8.67, 8.7, 8.67, 8.65, 8.62, 8.6, 8.62, 8.65, 8.67, 8.7, 8.67, 8.65, 8.62, 8.6, 8.55, 8.5, 8.45, 8.4, 8.35, 8.3, 8.25, 8.2, 8.47, 8.75], 'pv_power_forecast': [418.4, 552.9, 707.5, 862.1, 1019.2, 1176.3, 1359.15, 1542.0, 1619.15, 1696.3, 1834.0, 1971.7, 2017.8, 2063.9, 1963.2, 1862.5, 1817.55, 1772.6, 1797.75, 1822.9, 1807.8, 1792.7, 1737.9, 1683.1, 1602.35, 1521.6, 1404.85, 1288.1, 1144.85, 1001.6, 843.85, 686.1, 523.95, 361.8, 269.8, 177.8, 120.8, 63.8, 34.8, 5.8, 2.9, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 6.05, 12.1, 57.95, 103.8, 184.75, 265.7]}

here is my runtime payload

      {% macro time_to_timestep(time) -%}
        {{ ( max(0, ((today_at(time) - now().replace(minute=0, second=0, microsecond=0)) / timedelta(minutes=15))) | round(0, 'ceiling')) % 96 }}
      {%- endmacro %}
      {% macro time_to_timestep_tomorrow(time) -%}
        {{ ( (((today_at(time) + timedelta(days=1) - now().replace(minute=0, second=0, microsecond=0)) / timedelta(minutes=15))) | round(0, 'ceiling')) }}
      {%- endmacro %}
      {% macro time_to_timestep2(time) -%}
        {%- set target_time = today_at(time) -%}
        {%- if target_time < now() -%}
          {%- set target_time = target_time + timedelta(days=1) -%}
        {%- endif -%}
        {%- set quarterhours_until = (target_time - now()).total_seconds() / 900 -%}
        {{ max(0, quarterhours_until) | round(0, 'ceiling') }}
      {%- endmacro %}
      
      {%- set horizon = min(96, (state_attr('sensor.electricity_price', 'future') | map(attribute='total') | list) | length)  -%}
      {% set horizon = 96 %}
      
      {%- set temperatures = (state_attr("sensor.emhass_oat_forecast_15min", "data") | map(attribute="temperature") | list)[:horizon] -%}
      {%- set avg_temp = (temperatures | sum) / temperatures | length -%}
      {%- set heating_power = max(450, (((15 - avg_temp) * 110) - 200)) -%}
      
      {% set schalter0 = states('input_boolean.emhass_hotwater') %}
      {% set schalter1 = states('input_boolean.emhass_dishwasher') %}
      {% set schalter2 = states('input_boolean.emhass_washingmaschine') %}
      {% set schalter4 = states('input_boolean.emhass_kochen') %}
    


      
      {% set current_state = [
          states('select.heizung_strom') == "Hot Water",
          states('input_boolean.dishwasher') == "on",
          states('binary_sensor.washing_maschine') == "on",
          states('select.heizung_strom') == "Heating",
          states('sensor.pro3em_herd_total_active_power')|float(0) > 10.0
      ] %}
      
      {% set start_timesteps = [
          0,
          0,
          0,
          0,
          time_to_timestep2('18:00')|int
      ] %}

      {% set end_timesteps = [
          time_to_timestep2('15:00')|int,
          time_to_timestep2(states('input_datetime.emhass_deferrable3_finish_time'))|int - 2,
          0,
          0,
          time_to_timestep2('19:00')|int
      ] %}
      
      
      {% set wert0 = 0 %}
      {% set wert1 = min(2, end_timesteps[1]/4) %}
      {% set wert2 = states('input_number.emhass_washingmaschine_t')|int %}
      {% set wert4 = 1 %}
    
      
      {%- set operating_hours = [
        wert0 if schalter0 == 'on' else 0,
        wert1 if schalter1 == 'on' else 0,
        wert2 if schalter2 == 'on' else 0,
        0,
        wert4 if schalter4 == 'on' else 0
      ] -%}
      
      {% set nominal_power = [
          1500,
          300,
          states('input_number.emhass_washingmaschine_p')|round(0),
          heating_power,
          1000
      ] %}
      
      {% set minimum_power = [
          1400,
          0,
          0,
          440,
          0
      ] %}
      
      {% set start_penalty = [
          100,
          5, 
          1,
          100,
          1
      ] %}
      
      
      {% set def0_temperature_setpoint = 45.0 %}
      {% set def0_temperature_reduced_setpoint = 32.0 %}
      {% set def0_overshoot_temperature = 48.0 %}
      
      {%- set def0_heated_intervals = [[time_to_timestep("17:00")|int, time_to_timestep("18:00")|int], [time_to_timestep_tomorrow("17:00")|int, time_to_timestep_tomorrow("18:00")|int]] -%}
      
      
      {% set t_outdoor = 10.0 %}
      {% set t_supply_min = 23.0 %}
      {% set def3_temperature_setpoint = t_supply_min + (20-t_outdoor)*0.33 %}
      {% set def3_temperature_setpoint = 21.5 %}
      {% set def3_temperature_reduced_setpoint = 21.0 %}
      {% set def3_overshoot_temperature = 26.0 %}
      
      {%- set def3_heated_intervals = [[time_to_timestep("06:00")|int, time_to_timestep("22:00")|int], [time_to_timestep_tomorrow("06:00")|int, time_to_timestep_tomorrow("22:00")|int]] -%}
      
      
      {% set prices = (state_attr('sensor.electricity_price', 'future') | map(attribute='total') | list) %}
      {% set last = max(0.35, (prices[-1] if prices|length > 0 else 0)) %}
      {% set price_filled = prices + ([last] * (horizon - prices|length)) %}

      
      {%- set pv_power_forecast = namespace(all=[]) -%}
      {% for i in range(horizon) %}
        {%- set pv_power_forecast.all = pv_power_forecast.all + [ 0.0 ] -%}
      {% endfor %}
      {%- set load_power_forecast = namespace(all=[]) -%}
      {% for i in range(horizon) %}
        {%- set load_power_forecast.all = load_power_forecast.all + [ 0.0 ] -%}
      {% endfor %}
      {
        
        "method": "mlforecaster",
        "prediction_horizon": {{ horizon }},
        "entity_save": true, 
        "publish_prefix":"emhass_",
        "soc_init":{{states('sensor.byd_battery_box_premium_hv_ladezustand')| float(0)/100}},
        "soc_final":0.5,
        "load_cost_forecast": {{ price_filled[:horizon] | tojson }},
        "operating_hours_of_each_deferrable_load": {{ operating_hours }},
        "def_current_state": {{ current_state | tojson }},
        "def_start_penalty": {{ start_penalty | tojson }},
        "start_timesteps_of_each_deferrable_load": {{ start_timesteps | tojson }},
        "end_timesteps_of_each_deferrable_load": {{ end_timesteps | tojson }},
        "nominal_power_of_deferrable_loads": {{ nominal_power | tojson }},
        "minimum_power_of_deferrable_loads": {{ minimum_power | tojson }},
        "def_load_config": [
          {
            {% if schalter0  %}
            "thermal_config": {
            "heating_rate": 15,
            "cooling_constant": 0.004,
            "overshoot_temperature": {{def0_overshoot_temperature}},
            "start_temperature": {{  states('sensor.bt6_hw_load_40014') | float(40.0) }},
            "desired_temperatures": [
              {%- set comma = joiner(", ") -%}
              {%- for i in range(horizon) -%}
                {%- set timestep = i -%}
                {{ comma() }}
                {% for interval in def0_heated_intervals if timestep >= interval[0] and timestep < interval[1] %}
                {{ def0_temperature_setpoint }}
                {%- else -%}
                {{ def0_temperature_reduced_setpoint }}
                {%- endfor %}
              {%- endfor %}
            ]}
            {% endif %}
          },
          {},
          {},
          {
          {% if states('switch.allow_heating_47371') == "on"  %}
            "thermal_config": {
            "heating_rate": 5,
            "cooling_constant": 0.018,
            "overshoot_temperature": {{def3_overshoot_temperature}},
            "start_temperature": {{  states('sensor.bt2_supply_temp_s1_40008') | float(21.0) }},
            "desired_temperatures": [
              {%- set comma = joiner(", ") -%}
              {%- for i in range(horizon) -%}
                {%- set timestep = i -%}
                {{ comma() }}
                {% for interval in def3_heated_intervals if timestep >= interval[0] and timestep < interval[1] %}
                {{ def3_temperature_setpoint }}
                {%- else -%}
                {{ def3_temperature_reduced_setpoint }}
                {%- endfor %}
              {%- endfor %}
            ]}
          {% endif %}
          },
          {}
        ],
        "outdoor_temperature_forecast": {{ temperatures | tojson() }},
        "pv_power_forecast": {{ state_attr('sensor.emhass_calculated_solcast', 'data')[:horizon] | map(attribute='pv_estimate10') | list | tojson() }}
      }