Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
May 6, 2026AerospaceOpen Access

A Component-Decoupled and Physics-Constrained Hybrid Modeling Framework for Turbojet Engine Performance Prediction

View Full Paper
Ask AI
Bookmark
Share

Authors

顾顾怀平Northwestern Polytechnical UniversityLJLinyuan JiaNorthwestern Polytechnical UniversityHDHui DuanAviation Industry Corporation of China (China)

Discussion

Loading...

Member takes

Overview

A hybrid modeling approach improves performance prediction accuracy in turbojet engines, suggesting better condition monitoring and health management.

Key Points

  • The research aims to create a hybrid modeling framework for turbojet engine performance prediction.
  • Developed a component-decoupled, physically constrained modeling framework.
  • Applied physics-guided feature engineering and mutual-information-based feature selection.
  • Coupled predictions via aerothermodynamic constraints to reconstruct parameters.
  • Achieved 1.157% maximum relative deviation and 0.226% average relative deviation for thrust.
  • Model predictions for key gas path parameters showed deviations within 0.3%.
  • Demonstrated improvement in accuracy and generalization compared to data-driven models.

Cite This Study

顾怀平 et al. (2026) studied this question.

synapsesocial.com/papers/69fa980604f884e66b531dcchttps://doi.org/10.3390/aerospace13050425
View Full Paper
Ask AI
Bookmark
Share