The pre-existing efficiency formula `η = flow/power` produced tiny SI-unit
values (m³/J ≈ 1e-5), was monotonic in ctrl for centrifugal-pump curves
(no interior peak), and made NCog collapse to 0 — which cascaded into MGC
reporting BEP-position 0.0% always. Replaced with hydraulic efficiency
η = (Q·ΔP)/P_shaft, the dimensionless 0..1 ratio that has a real BEP and
matches the form MGC's group-level math uses.
- prediction/efficiencyMath.js:
* calcEfficiencyCurve takes pressureDiffPa; η = 0 when dP missing
* calcCog guards (yMax > yMin) before computing NCog (was unguarded /0)
* calcEfficiency falls back to predictFlow.currentF when measured ΔP is
missing, so predicted-variant calls still produce a meaningful η before
the differential measurement settles
- specificClass.js:
* Asset-registry lookup renamed: 'machine' → 'rotatingmachine' (matches
the datasets/assetData/ rename in generalFunctions). The error path
quotes the new filename so operators can find it.
* Two-call-site fix: with default-param stateConfig={}, the single-arg
constructor path (BaseNodeAdapter calls `new Machine(this.config)`
after pre-setting Machine._pendingExtras) was silently clobbering the
pre-set extras. Only overwrite when the caller explicitly passes them.
* Push port 0 deltas (notifyOutputChanged) after prediction updates so
dashboards see state + predicted-flow changes as they happen.
- pressure/pressureRouter.js: routing + fallback hardening (the trigger
for the bep-distance-cascade reproduction).
- display/workingCurves.js: Q-H curve generator extended.
- New tests:
* test/integration/qh-curve.integration.test.js — Q-H curve shape
* test/integration/bep-distance-cascade.integration.test.js — reproduces
the dashboard report (absDistFromPeak=0, NCog=0, efficiency=0 after a
setpoint move) at the unit level so future regressions fail loudly.
Full suite: 214/214 pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
specificClass._setupCurves now calls assetResolver.resolveAssetMetadata
to derive supplier/type/units from the model id, instead of trusting
denormalized fields on the node config. If the model isn't in the
registry, installs a null-predictor stub and logs a clear "pick a model
from the asset menu" error rather than crashing.
rotatingMachine.html: defaults block trimmed (supplier/category/assetType
were stale copies of registry data).
Tests:
- New test/basic/assetMetadata.basic.test.js covers the registry-resolve
path and the missing-model fallback.
- nodeClass-config / error-paths / nodeClass-routing / factories /
abort-deadlock fixtures updated to the trimmed asset shape.
- 209/209 tests pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Same as MGC — UnitPolicy property bags replace the manual _unitView/
unitPolicyView reassignment. specificClass.js 400→377. 196/196 tests
still pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When PS commanded turnOffAllMachines, executeSequence's interruptible
abort path triggered transitionToState('operational'), which auto-picked
up the queued delayedMove and re-started the pump. Pump bounced
accelerating ↔ decelerating forever and never reached idle.
Clear state.delayedMove at the top of shutdown/emergencystop sequences
so a user-commanded stop cancels any pending move.
Observed live: in pumpingstation-complete-example the basin drained
past stopLevel and equilibrated at ~0.3 m with one pump stuck at min
flow. With this fix pumps shut down cleanly at stopLevel.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds a parallel set of Predict instances (groupPredictFlow / Power / Ctrl)
that share input curves with the pump's individual predicts but maintain
their own operating point. MGC drives these via setGroupOperatingPoint()
to evaluate every pump curve at one shared manifold differential during
combination optimization, without corrupting each pump's own diagnostic
outputs (which track that pump's local sensors).
Created lazily on first use so pumps without an MGC parent pay nothing.
Pairs with generalFunctions Predict.shareInputsFrom plumbing.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- rotatingMachine.html: add default name:{value:""} to the editor
defaults block (standard Node-RED pattern; was missing).
- nodeClass.js: clear node status badge on close — matches the
pattern already in other EVOLV node close handlers.
- specificClass.js: remove the `(x <= 0) ? 0 : ...` guard in the
flow and power prediction methods. The guard was redundant:
predictions only run while the FSM is in an active state
(operational / starting / warmingup / accelerating / decelerating),
none of which produce x=0. Math.max(0, rawFlow) still clamps
negative extrapolation. Net: same behaviour in production, less
dead code.
All 10 basic tests pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The bogus machineCurve default at pressure "1" (fixed in generalFunctions
086e5fe) made fValues.min=1, which let sub-curve differentials pass
unclamped. With the fix, fValues.min=70000 (the real curve minimum) and
low differentials get clamped. Three tests that accidentally depended on
the bogus min=1 behavior are updated:
- coolprop test: expects fDimension clamped to curve minimum when
differential < curve range
- pressure-initialization test: uses pressures whose differential falls
WITHIN the curve range (900 mbar = 90000 Pa > 70000 Pa minimum)
- sequences test: tests upper-bound constraint with setpoint > max,
then confirms a valid setpoint is applied as-is (was incorrectly
asserting any setpoint would be clamped to max)
Trace instrumentation from debugging session removed.
91/91 tests green.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
At ctrl=0% with high backpressure, the curve prediction extrapolates to
large negative values (backflow through a stopped pump). This produced
confusing chart readings (-200+ m³/h for an idle pump) and polluted
downstream consumers like MGC efficiency calculations.
Fix: in both calcFlow and calcPower, if the controller position x ≤ 0
the prediction is clamped to 0 regardless of what the spline returns.
For x > 0, predictions are also clamped to ≥ 0 (negative flow/power
from a running pump is physically implausible for a centrifugal machine).
91/91 tests still green — no existing test asserted on negative
flow/power values at ctrl=0.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Runtime:
- executeSequence now normalizes sequenceName to lowercase so parent
orchestrators that use 'emergencyStop' (capital S) route correctly to
the 'emergencystop' sequence key. Closes the "Sequence 'emergencyStop'
not defined" warn seen when commands reach the node during accelerating.
- When a shutdown or emergencystop sequence is requested while the FSM is
in accelerating/decelerating, the active movement is aborted via
state.abortCurrentMovement() and the sequence waits (up to 2s) for the
FSM to return to 'operational' before proceeding. New helper
_waitForOperational listens on the state emitter for the transition.
- Single-side pressure warning: fix "acurate" typo and make the message
actionable.
Tests (+15, now 91/91 passing):
- test/integration/interruptible-movement.integration.test.js (+3):
shutdown during accelerating -> idle; emergencystop during accelerating
-> off; mixed-case sequence-name normalization.
- test/integration/curve-prediction.integration.test.js (+12):
parametrized across both shipped pump curves (hidrostal-H05K-S03R and
hidrostal-C5-D03R-SHN1). Verifies loader integrity, mid-range prediction
sanity, flow monotonicity in ctrl, inverse-pressure monotonicity, CoG
finiteness, and reverse-predictor round-trip.
E2E:
- test/e2e/curve-prediction-benchmark.py: live Dockerized Node-RED
benchmark that deploys one rotatingMachine per curve and runs a per-pump
(pressure x ctrl) sweep inside each curve's envelope. Reports envelope
compliance and monotonicity.
- test/e2e/README.md documents the benchmark and a known limitation:
pressure below the curve's minimum slice extrapolates wildly
(defended by upstream measurement-node clamping in production).
UX:
- rotatingMachine.html: added placeholders and descriptions for Reaction
Speed / Startup / Warmup / Shutdown / Cooldown. Expanded the Node-RED
help panel with a topic reference, port documentation, state diagram,
and prediction rules.
Docs:
- README.md rewritten (was a single line) with install, quick start,
topic/port reference, state machine, predictions, testing, production
status.
Depends on generalFunctions commit 75d16c6 (state.js abort recovery and
rotatingMachine schema additions).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Safety:
- Async input handler: await all handleInput() calls, prevents unhandled rejections
- Fix emergencyStop case mismatch: "emergencyStop" → "emergencystop" matching config
- Implement showCoG() method (was routing to undefined)
- Null guards on 6 methods for missing curve data
- Editor menu polling timeout (5s max)
- Listener cleanup on node close (child measurements + state emitter)
- Tick loop race condition: track startup timeout, clear on close
Prediction accuracy:
- Remove efficiency rounding that destroyed signal in canonical units
- Fix calcEfficiency variant: hydraulic power reads from correct variant
- Guard efficiency calculations against negative/zero values
- Division-by-zero protection in calcRelativeDistanceFromPeak
- Curve data anomaly detection (cross-pressure median-y ratio check)
- calcEfficiencyCurve O(n²) → O(n) with running min
- updateCurve bootstraps predictors when they were null
Tests: 43 new tests (76 total) covering emergency stop, shutdown/maintenance
sequences, efficiency/CoG, movement lifecycle, output format, null guards,
and listener cleanup.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace hardcoded position strings with POSITIONS.* constants.
Prefix unused variables with _ to resolve no-unused-vars warnings.
Fix no-prototype-builtins where applicable.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
