Prediction of <scp> SO <sub>2</sub> </scp> concentration in <scp>WFGD</scp> system based on <scp>GWO</scp> optimized <scp>CNN</scp> ‐ <scp>BiLSTM</scp> ‐attention
Bibliographic record
Abstract
Abstract Due to external disturbances, the parameters of the desulphurization system are uncertain, and their relationships are complex, which makes it difficult to predict the concentration of SO 2 at the desulphurization system outlet. In this paper, grey wolf optimization (GWO) optimized convolutional neural network (CNN)‐bi‐directional long short‐term memory (BiLSTM)‐Attention algorithm was used for prediction, and the problem of low SO 2 concentration prediction accuracy at outlet has been resolved. First, the outliers of the thermal power plant desulphurization data were processed using the local outlier factor (LOF) algorithm. Secondly, CNN‐BiLSTM model was constructed using CNN and BiLSTM, and attention module was added to realize feature extraction and better capture the regularity of input data. Then, the CNN‐BiLSTM‐Attention model was optimized using GWO and its hyperparameters were improved. Finally, based on the Matlab R2023a platform, the prediction comparison as well as the error analysis of the desulphurization data were carried out. In the prediction of SO 2 concentration in low‐flow continuous slurry supply mode, the error of the combined model decreased by 23.2% on average compared to the CNN‐BiLSTM‐Attention model. In the prediction of SO 2 concentration in the high‐flow intermittent slurry supply mode, the error of the combined model decreased by 20.8% on average. According to the results, the combined model surpasses both the single model and several other algorithmic combination models in terms of performance metrics, and the predictions are more accurate.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".