Document sensor data we want logged that is not currently present #7

Closed
opened 2026-09-09 03:14:46 +00:00 by agent-pi · 1 comment
Collaborator

The analysis work keeps hitting the same walls: battery flow is inferred as a residual, the SoC 95-99 taper is excluded instead of modelled, and the irradiance-to-power model needs per-season slopes because the sensor is horizontal while the arrays are tilted. The sensor data that would fix these is not collected anywhere in the current log.

These wishes should be written down somewhere durable so instrumentation decisions (and any new MQTT topics) have a reference, with the analysis each field would unlock.

The analysis work keeps hitting the same walls: battery flow is inferred as a residual, the SoC 95-99 taper is excluded instead of modelled, and the irradiance-to-power model needs per-season slopes because the sensor is horizontal while the arrays are tilted. The sensor data that would fix these is not collected anywhere in the current log. These wishes should be written down somewhere durable so instrumentation decisions (and any new MQTT topics) have a reference, with the analysis each field would unlock.
Author
Collaborator

Plan of attack

  • Add a top-level Desired sensor data (not yet logged) section to mqtt-topic-logger.md (the logger reference doc, where TOPICS config lives), listing each missing field, its likely source, and the analysis it unlocks - grounded in existing findings (residual-based fast-charge drain, SoC 95-99 taper exclusion, seasonal slope split R2 0.69).
  • Also note what is already covered (zappi mode/plug/cooldown, forecast snapshots - verified ~100% presence since 2026-08-10) so the list stays honest.
  • Fields whose availability depends on hardware/BMS support are phrased conditionally, not as confirmed capabilities.
  • Commit :memo: document desired sensor data not yet in the log on fix/issue-5-analyze-event-lines so it lands with PR #6.
**Plan of attack** - Add a top-level **Desired sensor data (not yet logged)** section to `mqtt-topic-logger.md` (the logger reference doc, where TOPICS config lives), listing each missing field, its likely source, and the analysis it unlocks - grounded in existing findings (residual-based fast-charge drain, SoC 95-99 taper exclusion, seasonal slope split R2 0.69). - Also note what is already covered (zappi mode/plug/cooldown, forecast snapshots - verified ~100% presence since 2026-08-10) so the list stays honest. - Fields whose availability depends on hardware/BMS support are phrased conditionally, not as confirmed capabilities. - Commit `:memo: document desired sensor data not yet in the log` on `fix/issue-5-analyze-event-lines` so it lands with PR #6.
Sign in to join this conversation.
No labels
No milestone
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Reference
marvin8/IrradiationToPowerProduction#7
No description provided.