Gas Turbine Thermal Digital Twin: Developing a Data Engineering Framework
Bibliographic record
Abstract
Abstract A gas turbine engine (GT) is a complex power-producing machine that operates under high speed, pressure and temperature. Monitoring its health in real time is essential for ensuring safety and reliability. However harsh operating conditions, limited accessibility and high instrumentation cost, make it challenging to get sensor data from all life limited locations. This can be addressed by developing and deploying a Digital Twin (DT) which utilizes engine data from tests, fleet operations and numerical analysis, to calculate local and bulk metal temperatures at locations of interest. Development of this Digital Twin needs a specialized framework for data collection, pre-processing, model training, and evaluation. This paper focuses on data acquisition, transformation and engineering steps that are performed prior to the development of a DT, which are critical for enhancing its accuracy and robustness. Data acquisition involves fetching real-time sensor data from proprietary cloud system. Data transformation involves converting this data into a format readable by Python. The dataset includes a wide range of operating conditions that a GT experiences throughout its lifecycle, providing an exhaustive dataset for training the DT model. Data engineering involves removal of false sensor readings, filtration of dataset, data imputation for missing data, and noise reduction to deal with occasional spikes and noise, normalization of input features and target(s). These operations have been performed using Python built-in standard libraries like Pandas, NumPy, scikit-learn. In addition, statistical techniques like Pearson’s correlation are utilized in the feature engineering process. The paper is intended to illustrate the process to develop a data engineering framework to develop GT Digital Twin for metal temperature prediction.
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 | 0.007 | 0.016 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.005 | 0.007 |
| Open science | 0.004 | 0.007 |
| Research integrity | 0.001 | 0.004 |
| Insufficient payload (model declined to judge) | 0.006 | 0.004 |
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".