Implement auto populate multi switch entities

[bramstroker]

No description provided.

[michalk-k]

@bramstroker I went through changes to find out how should the profile configuration for such a multiswitch looks like and how it works. Kudos to you for updating the documentation.

I found it in docs but it's not exactly explained how to interpret that:

"multi_switch_config": {
    "power": 0.8,
    "power_off": 0.25
  },

Could you elaborate, please?
I have measured Shelly 2.5.
It's consumption is:

• 0.01W constant self-consumption
• 0.00325W for each relay turned on. It has two relays.

So with one relay on the total consumption is 0.01 + 0.00325 = 0.01325W, for both relays on = 0.0165W

How should it be reflected in the new configuration section? how does it correlate with standby_power profile attribute?

[bramstroker]

Also have a look here:
https://docs.powercalc.nl/strategies/multi-switch/

You might try:

"standby_power": 0.01,
"multi_switch_config": {
    "power": 0.00325,
    "power_off": 0
  },

Or:

"multi_switch_config": {
    "power": 0.00325,
    "power_off": 0.005
  },

I think I prefer the first solution, but I did not test this configuration yet.

[bramstroker]

But I don't think your figures are correct, as all the shelly's at least consume 0.5W in standby.
As it are WiFi devices and typically these radio's consume more than zigbee and zwave.
0.5 at least.
Relay's typically also consume 0.3+ as long as it are no solid state relays, which I don't think the shelly 2.5 are.

[michalk-k]

I would also prefer the first configuration. If both gives the same result, the first one is more self-explanatory.

Regarding the consumption of Shelly 2.5, thanks for the heads-up. Of course, those are bad numbers. I gave you current, instead of power. Just copy-paste error.

[michalk-k]

@bramstroker let's update S2.5 values and discuss something.

So, standby power is 0.677W, then 0.324W per relay.
I assume we agreed on following representation:

"standby_power": 0.677,
"multi_switch_config": {
    "power": 0.324,
    "power_off": 0
  },

Just a note, that this is kind of not consistent with fixed and linear strategies, for example for Shelly 1PM I measured:

"calculation_strategy": "fixed",
  "standby_power": 0.739,
  "fixed_config": {
    "power": 1.0
  },

Note that "power" here represents a total measured power, while in the case of a multi-switch, we are putting the delta values there.
I can bet I saw some attribute that changes the behavior for fixed/linear to be more like multi switch. Anyway... it's something to keep in mind IMO.

[bramstroker]

I did some testing using the preferred configuration but found some issues down the road.
I have fixed them now with #2877. will issue a new version shortly.
Also made the power_off optional, as it's not necessary when defining standby_power for the whole device.

So the configuration will look like:

{
    "standby_power": 0.677,
    "multi_switch_config": {
        "power": 0.324
    }
}

Regarding the inconsistency / example for fixed you gave this is old configuration.
Please see the docs for all the up to date and correct configurations to use now when making profiles for the library:
https://docs.powercalc.nl/library/device-types/smart-switch/
https://docs.powercalc.nl/library/device-types/power-meter/

Notice the addition of "only_self_usage": true which indicates the profile is only used to provide self usage consumption figures.
This will make sure the user is not asked to provide power values for the connected appliance.
Also I plan to utilize this information in the future to add power to the all_standby_power sensor, as it's missing the self usage of devices which are turned on.

I will update the current profiles in the library to all match the configuration examples in the docs and make them all consistent.
But I can only do that after v1.17 stable hit the road, and is installed by most users.