Improvement
- feat: mask the demand/capacity charge to a tariff demand window (capacity_charge_window) (@hossamnagy)
- Preparing new release v0.18.0 (@davidusb-geek)
Documentation
- docs: add forecast-to-MPC study (@OEB11)
Fix
- fix: size the mlforecaster load forecast history retrieval from the trained model's lags (#1067) (@LesIT1)
- fix: tighten the battery identification RTE guardrail to reject near-lossless fits (@LesIT1)
- Fix: duplicate forecast_dates when Forecast is constructed on an exact half-interval tie second (@LesIT1)
- fix: battery identification no longer counts throughput across recorder gaps (#1051) (@LesIT1)
- fix: keep the relaxed-LP rescue one-shot so it cannot poison the cached problem (#1048) (@LesIT1)
What's Changed
- fix: keep the relaxed-LP rescue one-shot so it cannot poison the cached problem (#1048) by @LesIT1 in #1059
- docs: add forecast-to-MPC study by @OEB11 in #1062
- fix: battery identification no longer counts throughput across recorder gaps (#1051) by @LesIT1 in #1063
- Fix: duplicate forecast_dates when Forecast is constructed on an exact half-interval tie second by @LesIT1 in #1064
- feat: mask the demand/capacity charge to a tariff demand window (capacity_charge_window) by @hossamnagy in #1066
- fix: tighten the battery identification RTE guardrail to reject near-lossless fits by @LesIT1 in #1070
- fix: size the mlforecaster load forecast history retrieval from the trained model's lags (#1067) by @LesIT1 in #1071
New Contributors
Full Changelog: v0.18.0...v0.18.1