Intelligent calibration method for coal mine underground methane sensor
By optimizing the standard concentration trajectory and information matrix driven by metrological certificates, combined with robust parameter estimation and Wasserstein centroid optimization, the drift and consistency problems of downhole methane sensors were solved, enabling robust online calibration and traceable monitoring, and reducing the difficulty of false alarms, missed alarms, and replays.
Patent Information
- Application Number
- CN202511322226.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-16
- Publication Date
- 2026-05-08
- Estimated Expiration
- 2045-09-16
AI Technical Summary
In coal mine environments, underground methane sensors suffer from sensitivity and time constant drift, inconsistent results at multiple points, and a lack of traceable on-site calibration capabilities. This leads to misjudgment of monitoring network thresholds, false alarms and missed alarms, difficulties in accident review, and high maintenance costs.
A standard concentration trajectory driven by metrological certificates is adopted, combined with information matrix and H-norm optimization, and robust online correction is achieved through robust parameter estimation and Wasserstein centroid optimization. Parameter identification and correction are performed using near-field communication sessions, acoustic pulse sealing diagnostics, and robust deconvolution techniques.
It enables the traceability and accountability of standard concentration trajectories, explicitly models leakage uncertainty, enhances information gain, ensures consistent correction across devices, and reduces the difficulty and maintenance cost of accident review.