Sensor fault handling method and apparatus for aeroengine multi-sensor
By employing a dual-task learning framework combining an aero-engine sensor fault detection network and an isolation network, abnormal sensors are identified and isolated, and target sensing signals are generated. This solves the problem of low accuracy in sensor fault diagnosis in existing technologies and ensures stable engine operation.
CN120316609BActive Publication Date: 2026-06-12TSINGHUA UNIVERSITY
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- TSINGHUA UNIVERSITY
- Filing Date
- 2025-02-20
- Publication Date
- 2026-06-12
Smart Images
Figure CN120316609B_ABST
Abstract
The application relates to an aero-engine multi-sensor sensor fault processing method and device. The method comprises the following steps: acquiring sensor input data of multiple sensors, and generating reconstructed sensor data of each sensor through a fault detection network based on the sensor input data of each sensor; identifying an abnormal sensor based on the sensor input data of each sensor and the reconstructed sensor data of each sensor, and starting a data isolation process of the abnormal sensor through an isolation network of the abnormal sensor; predicting a target sensor signal of the abnormal sensor through the isolation network based on the sensor input data of each sensor, and sending the target sensor signal to a control center of an aero-engine. The method can improve the accuracy of sensor fault diagnosis and processing.
Need to check novelty before this filing date? Find Prior Art