The complete reasoning chain
Inputs with physical meaning
- Digester and ambient temperature
- pH and organic loading rate
- Hydraulic retention time
- Total and volatile solids
- C/N ratio and moisture
- Prior methane observations
- Reactor identity and time
Inputs GFIS must not invent
- Missing methane targets are never imputed.
- Total biogas is never renamed as methane.
- Validation/test observations never fit scaling.
- Synthetic perturbations are labelled synthetic.
- Reactor identity is retained across sequences.
Current hybrid engine
XGBoost branch
Learns nonlinear interactions among present process variables and past-only lag/rolling features. Tree structure can be inspected through feature importance and SHAP in the final explainability layer.
LSTM neural branch
Reads an ordered window of reactor observations and retains temporal state through input, forget and output gates. Every epoch, seed, window and checkpoint is stored.
Physics/stability branch
Audits methane feasibility against a VS-dependent ceiling and interprets VFA/ALK as Stable, Warning or Critical. The current ceiling is post-prediction; a genuine loss-term ablation is the next engine upgrade.
The ensemble weight α must be selected using validation data—not the final test period. In the public DS-03 experiment, the LSTM was not promoted merely because one final-period score looked favourable.
Inside the neural engine
A neural engine already exists: the PyTorch LSTM. Version 3 makes its state update explicit and defines the next physics-guided neural engine as an evidence-controlled extension, not an unexplained new model.
ft = σ(Wxfxt + Whfht−1 + bf)
ct = ft⊙ct−1 + it⊙tanh(Wxgxt + Whght−1 + bg)
ŷt = wyTht + by
Existing neural engine
One-layer LSTM, hidden size 32, chronological windows 3/7/14/28, repeated seeds, early stopping, saved epochs and checkpoints. It is real and executable, but DS-03 evidence does not support longer memory as champion.
Proposed physics-guided residual neural engine
A compact recurrent network predicts a correction around a mechanistic or empirical baseline. The correction is penalized when it violates methane-yield bounds or inconsistent stability states. This must be implemented as an ablation-controlled experiment before being claimed as a result.
Each λ will be configuration-controlled. The report will show the accuracy–consistency trade-off, violation rate and calibration—not only the best error score.
How GFIS removes black-box behaviour
| Black-box risk | GFIS transparency control | Visible evidence |
|---|---|---|
| Unknown data origin | Immutable archive, licence, citation and SHA-256 | Dataset catalogue and run manifest |
| Future leakage | Chronological splits and fold-local preprocessing | Split manifest and rolling folds |
| Unexplained temporal memory | Sequence windows, seed repetition and epoch histories | Epoch CSV, checkpoints and window study |
| Physically impossible methane | Raw prediction, upper bound, violation flag and correction retained | Simulator trace and physics-ablation plan |
| Opaque alarm | VFA/ALK value, thresholds and stability state shown together | Warning rows and simulator memory |
| Single favourable metric | Persistence baseline, rolling folds and dispersion | Fold metrics and scientific decision |