Researchers have developed an automated workflow to generate a large, hybrid dataset for deep anomaly detection in chemical processes. This new dataset combines experimental data with simulated data, created using a novel Python-based simulator that employs differential-algebraic equations. The simulation approach accurately predicts experimental dynamics after calibration, enabling the consistent generation of time-series data for both normal operations and various anomalies. This hybrid dataset is now openly available and aims to facilitate research in simulation-to-experiment transfer and deep anomaly detection methods. AI
IMPACT Provides a unique dataset for advancing deep anomaly detection methods in chemical process monitoring.
RANK_REASON This is a research paper detailing a new dataset and simulation methodology for anomaly detection. [lever_c_demoted from research: ic=1 ai=1.0]
- arXiv
- batch distillation
- Deep Anomaly Detection for CNC Machine Cutting Tool Using Spindle Current Signals
- Hugging Face
- Justus Arweiler
- Python
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
