Open nicorusti opened 2 weeks ago
@alistair23 would it be ok to contribute your work to the main repository?
Fine with me
@alistair23 I have been running this for a couple of days, and I notice an issue with your control algorithm that is using instant power consumption to estimate remaining power for the day ( introduced in 0220d251123bcf6010643aa653c75326fb49db1a) :
Basically this triggers the appliance to turn on as soon as there is a sudden consumption peak, that causes remaining_power
to go down, while avg_excess_power
stays positive. Then appliance stays on as soon as it can be powered in with some solar energy, and then turn off as soon as there is another peak:
I might try using a better metric, maybe average daily power consumption, or average from the last x hours. Let me know if you see the same issue and if you have some ideas
I might try using a better metric, maybe average daily power consumption, or average from the last x hours. Let me know if you see the same issue and if you have some ideas
Ah, good point. An average of the last x hours seems like the best bet
Addresses #32 , based on work by @alistair23.
Some work done:
@alistair23 would it be ok to contribute your work to the main repository?
Pending issues:
Improvements: