A Data-Model Fusion Framework for Aero-Engine Degradation Diagnosis under Measurement Underdetermined Conditions

Abstract In aircraft engine gas-path analysis, the number of measurable parameters is often limited and smaller than the number of health parameters to be estimated, resulting in an underdetermined diagnostic problem that compromises the accuracy and reliability of engine degradation assessment. To address this issue, this paper develops a novel gas-path degradation diagnostic framework that integrates data-driven and model-based approaches, and proposes a new Hybrid Gas-Path Analysis with Data-Driven Priors (HGPA-DDP) method for engine degradation assessment under underdetermined conditions. First, a data-driven model is constructed to perform preliminary diagnostics on measurable parameters, generating prior information on component degradations. Then, this prior information is incorporated into a nonlinear gas-path analysis framework through Tikhonov regularization, enabling iterative optimization under physical-model constraints and achieving an effective fusion of physical mechanisms and data-driven advantages. The proposed method is validated through turbofan engine degradation simulations and on-wing operational data. The results demonstrate that, compared with conventional nonlinear gas-path analysis methods, the proposed HGPA-DDP approach significantly improves the accuracy and reliability of diagnostic results under underdetermined measurement conditions, providing a more reliable diagnostic tool for aircraft engine health management.