Superconductor quench detection method, system, and medium based on multi-sensor fusion

By combining multi-sensor fusion and LSTM networks, accurate prediction and timely intervention of superconductor quench risk are achieved, solving the problem that existing technologies cannot predict quench in advance, and improving the safety and reliability of superconductor operation.

CN121142418BActive Publication Date: 2026-06-26HEFEI CAS ION MEDICAL & TECHNICAL DEVICES CO LTD +2
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511339145.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-09-18
Publication Date
2026-06-26
Estimated Expiration
2045-09-18

AI Technical Summary

Technical Problem

Existing methods for detecting superconductor quenching rely on a single monitoring approach, which cannot fully capture complex changes, leading to misjudgments or missed detections. Furthermore, they cannot predict quenching risks in advance, resulting in time-consuming and costly post-event processing.

Method used

By employing multi-sensor fusion technology, multiple sensors are deployed to collect multi-dimensional raw data. LSTM networks are used to extract key features and predict the probability of failure, thereby assessing the risk level and executing intervention operations.

Benefits of technology

This enables early prediction and accurate assessment of the risk of superconductor quenching, reducing misjudgments and losses, and ensuring the safety and reliability of superconductors.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121142418B_ABST
    Figure CN121142418B_ABST
Patent Text Reader

Abstract

The application discloses a kind of based on multisensor fusion's superconductor quench detection method, system and medium related to superconductor quench detection technical field, the method includes: by the multiple different types of sensors arranged, acquisition is used to describe the multidimensional original data of superconductor operating state in first set time range;Key features are extracted based on the original data;The key features are characterized in the time series features of superconductor operating state evolution trend in the first set time range;The key features are input to LSTM network for processing, obtain the predicted features in future second set time range, and output the quench probability of superconductor based on the predicted features;Based on the quench probability, quench risk level is judged, and based on the quench risk level, intervention operation is executed.The method can solve the problem that measures are taken after quench has occurred, and quench risk cannot be predicted in advance and intervened.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Superconducting magnet quenching detection method based on artificial intelligence

    CN114002630A

  • Superconducting and quenching detection system and method

    CN120595208A