Developing a Time-Efficient Model for Solid Oxide Fuel Cells Using Self-Supervised Convolutional Autoencoder and Stateful LSTM Network
Bibliographic record
Abstract
The development of an efficient dynamic model for solid oxide fuel cell (SOFC) systems is essential for control strategies, process optimization, and fault diagnostics - key steps toward maximizing power generation and extending their lifetime. Long-Short-Term-Memory (LSTM), a potent type of recurrent neural network, has proven to be adept at modeling intricate physical dynamic systems using extensive input-output data. They are particularly well-suited for capturing long-range dependencies and temporal patterns. Nevertheless, due to memory and computing constraints in practice, training an LSTM network to effectively learn long-term dependency is challenging, which stems from the sequential processing nature of LSTMs. In this study, a self-supervised convolutional autoencoder (AE) is developed to learn a concise yet informative temporal representation of historical input data. These compressed hidden sequences are then input into an LSTM model, which is trained using a truncated backpropagation through time (TBPTT) algorithm. Once trained, an identical stateful LSTM model is reconstructed using the learned parameters, enabling the LSTM to predict the voltage using only the most recent past input data, rather than a sequence of data over which the model was trained. The efficacy of the proposed framework is validated utilizing diverse experimental data collected from a lab-scale SOFC. The results indicate while the model properly identifies the underlying dynamics of the SOFC, it significantly reduces the training and runtime prediction computation costs of a traditional LSTM model by 38% and 42% respectively. This enhancement holds promise for improving real-time control and diagnostics of SOFC systems, ultimately contributing to their improved performance and reliability.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".