Research on Data Reconstruction Methods Based on Multiple Models of Aero-Engine Sensor

Sensors are crucial to an aero-engine. If they fail, the control system will not be able to accurately assess the engine's state, cannot take timely adjustments to ensure safe operation of the engine, and may even result in accidents. Therefore, establishing a reliable algorithm for reconstructing sensor data from an aero-engine is essential to ensure safe and efficient operation. This paper proposes a data reconstruction algorithm based on multiple models, mainly focused on the following three aspects: By utilizing the above methods, high quality reconstruction effects (reconstruction error less than 1%) can be achieved by using engine test data from a fleet for training and testing.