de dev branch heeft nu een experimentele implementatie van auto adaptive regeling. Ik moet dit nog een tijdje testen, en als er proefkonijnen zijn hoor ik dat ook graag. Verder vroeg ik de AI om release notes te schrijven op basis van de geschreven code. LOL zo glad als een aal, maar veel beter dan ik ooit zou kunnen.
--
Auto-Adaptive Control
The
Auto-Adaptive Control feature transforms your Ecodan heat pump into a smart, self-learning system. It intelligently adjusts the heat pump's operation to match the unique thermal properties of your home, maximizing both comfort and energy efficiency.
When enabled, this feature continuously monitors the room temperature and compares it to the setpoint. It then fine-tunes the heat pump's flow temperature to correct any persistent errors, ensuring your home is always perfectly comfortable using the minimum amount of energy.
The system is designed to be highly robust and accounts for real-world scenarios such as:
- System Saturation: It pauses the learning process if the heat pump is already running at 100% capacity.
- Defrost Cycles: It intelligently ignores temperature data during defrost cycles to avoid drawing incorrect conclusions.
- Short-Cycling: It works alongside the Short-Cycle Mitigation feature to ensure stable operation.
Persistent Learning
A significant advantage of this system is that all learned parameters, specifically the crucial
heating_curve_offset and
cooling_curve_offset, are automatically saved to the ESP32's flash memory.
Unlike the native Mitsubishi auto-adaptive function, which can lose its learned data after a power cycle, this implementation ensures that a reboot or power outage
does not cost time to re-learn. The system immediately resumes operation with its last known optimal settings, ensuring continuous efficiency.
Important Note on Operating Mode
For this feature to function correctly, the heat pump must be set to a
Fixed Flow Temperature mode for both heating and cooling. The Auto-Adaptive algorithm takes over the role of the built-in heating curve by dynamically adjusting this fixed setpoint.
Configuration Parameters
All parameters are adjustable in real-time from the Home Assistant interface.
| Parameter | Description | Guidance & Default |
|---|
| Auto-Adaptive: Control | Enables or disables the entire Auto-Adaptive feature. When disabled, the system will revert to using the standard fixed flow temperature setpoints. | Default: `On` |
| Auto-Adaptive: Heating System Type | Tunes the algorithm's behavior to match your system's thermal inertia (response time). | Default: Underfloor Heating UFH: For slow, high-inertia systems. UFH + Radiators: For hybrid systems. Radiators: For fast, low-inertia systems. |
| Auto-Adaptive: Heating Curve Slope | Determines how aggressively the flow temp rises as the outside temp drops. | Default: 1.5 Low (0.6-0.8) for well-insulated homes with UFH. High (1.2-1.6) for older homes with radiators. |
| Auto-Adaptive: Cooling Curve Slope | Determines how aggressively the flow temp drops as the outside temp rises. | Default: 1.2 Low (0.8-1.2) for homes with good sun protection. High (1.8-2.5) for homes with high solar gain. |
| Auto-Adaptive: Max. Heating Flow Temperature | Sets a hard safety limit for the flow temperature during heating to protect floors. | Default: 38.0°C |
| Auto-adaptive: Min. Cooling Flow Temperature | Sets a hard safety limit for the flow temperature during cooling to prevent condensation. | Default: 18.0°C |
| Auto-Adaptive: Room Temperature source | Selects the source for the room temperature reading used by the algorithm. | Default: Room Thermostat Room Thermostat: Uses the Ecodan thermostat. Rest API: Uses an external sensor from Home Assistant. |
Fine-Tuning Initial Values in YAML
While the system learns automatically, providing a good starting point in your YAML configuration is crucial for immediate efficiency. This is especially true for the
`heating_curve_offset`, which is the baseline for the heating curve.
| System Type | Recommended `initial_value` for `heating_curve_offset` |
|---|
| Underfloor Heating | `23.0` (More efficient for well-insulated homes) |
| Underfloor Heating + Radiators | `26.0` (A good intermediate value) |
| Radiators | `30.0` (A common baseline for radiator systems) |
https://github.com/gekkek...dev/docs/auto-adaptive.md
:strip_icc():strip_exif()/u/192109/zonnebloem.jpg?f=community)
:no_upscale():strip_icc():strip_exif()/f/image/jWLR7YBYnsn5KcaMigW1De6l.jpg?f=user_large)
:strip_exif()/f/image/lkEl6K6fFqjI0nXtVNkMTxhM.png?f=user_large)
:no_upscale():strip_icc():strip_exif()/f/image/VG5NNPDWio8llRCFeWKsRP80.jpg?f=user_large)
/u/241179/Gewelt.JPG?f=community)
:strip_exif()/f/image/DSjlp87undklWSU3zsMmhxkr.png?f=user_large)
Kan je voor de zekerheid de temperatuur van je esp rond dat tijdstip checken?
dus ws zal dit niet werken
:strip_icc():strip_exif()/u/23263/crop619cf05922b03_cropped.jpg?f=community)
:strip_exif()/f/image/NEQ0TFqtV3hLXHHNB1b26XmC.png?f=user_large)
:strip_exif()/f/image/sTDyDVuHkPnp5PomUByAGwwY.png?f=user_large)
:strip_exif()/f/image/6HgHeaPus5ZVgUJlqMSM3Bol.png?f=user_large)
![[Afbeelding]](https://tweakers.net/i/cWALAh4cm2fkGt9Y_vkK_iURxTk=/fit-in/4000x4000/filters:no_upscale():strip_exif()/f/image/0cKAglE487F65UdxWUpxHvcQ.png?f=user_large){kind=link}
/f/image/nVwpj3DeVJW8qlDWXzk6NmDI.png?f=fotoalbum_large){kind=link}