update docs for HVAC heat pumps #57
Conversation
| | $\gamma_1, \gamma_2$ | Parameter | Learning rate coefficients (income, age) | - | 2 × 1 | | ||
| | $\delta_1, \delta_2, \delta_3$ | Parameter | Comfort penalty coefficients (income, residents, climate) | $/°C·hr | 3 × 1 | | ||
|
|
||
| **Coefficient Calibration Notes:** |
There was a problem hiding this comment.
The bulletpoints below this rendered as a single paragraph, rather than a list, when I rendered the .md using just docs
| ### Household Characteristic Functions | ||
|
|
||
| **Income Factor:** | ||
| $$f_{AMI} = \left(\frac{AMI}{80\%}\right)^{0.6} - 1$$ |
There was a problem hiding this comment.
just curious, is ^0.6 a standard value, or just a guess?
There was a problem hiding this comment.
Just a guess! values would need to be fine tuned later ( the importat part is that its <1)
|
|
||
| - Consumer is already on TOU rates (given) and must choose between default and TOU-friendly HVAC operation schedules. | ||
|
|
||
| - Consumers learn schedule performance through direct experience rather than external information sources. |
There was a problem hiding this comment.
Just curious, how might we model a one-time external source like a mailer, an email, or a suggestion from a friend? I just it might be as simple as an override of the stay/switch decision, with some changed-your-mind factor [0-1] to capture the probability
There was a problem hiding this comment.
For now, just implied as something in the base exploration parameter epsilon (we can say that epsilon is the probability of being prompted by something external to change). However, network or information affects can be modelled in the same way as the other factors (additive term)
|
|
||
| Before the agent learning begins, we actually run OCHRE to produce complete annual building simulations for both schedule types to reduce computational load. This can help us find an approximate average monthly bill: | ||
|
|
||
| $$ |
There was a problem hiding this comment.
This aren't previous defined in the table.
This are functions of time, right? P[t], where [t] is each OCHRE simulation step (15 mins?)?
There was a problem hiding this comment.
P are prior initialization values so they should be an expected value-esque prior assumption on the V_tou and V_default. Will check over that table again!
closes #50