Multi-function composite data generation and PIMamba model for fault diagnosis in sucker-rod pumping wells
Bibliographic record
Abstract
ABSTRACT Petroleum is a critical energy resource in modern society, and its exploration and production are essential for meeting global energy demands. Dynamometer cards are important graphics that reflect the operational conditions of pumping wells, and their recognition is crucial for optimizing oil well production and diagnosing faults. With the development of deep learning, several automated methods based on deep learning have been proposed to analyze the specific working conditions of pumping wells from dynamometer cards. However, the sucker rod production system (SRPS) operates in a complex and variable environment, resulting in scarce effective samples and dynamometer card features that are sparse and informationally limited. To overcome these challenges, we propose a multi-function composite data generation paradigm that integrates diverse functional characteristics, generating 11 classes of highly interpretable single-condition images as training data for a prior model. This establishes a foundation of prior knowledge for training on subsequent actual condition data. Additionally, we introduce the Patch Importance Mamba (PIMamba) model, a dynamometer card recognition framework based on the State Space Model (SSM) architecture. The PIMamba model includes a Patch Importance (PI) module that assigns higher weights to data blocks containing key feature information, effectively filtering out irrelevant or low-sensitivity data and enhancing feature extraction precision and efficiency. In the Gaskule area of the western Qaidam Basin, PIMamba achieved a dynamometer card recognition accuracy of 94.73%, offering a novel approach to fault recognition in dynamometer cards and highlighting the significant potential of deep learning in the petroleum sector.
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 | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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".