A Machine Learning Framework for Turbofan Health Estimation via Inverse Problem Formulation

2026-04-09

AI Summary

Health state estimation for turbofan engines constitutes an ill-posed inverse problem, constrained by sparse sensor measurements and highly nonlinear thermodynamic characteristics. Existing research lacks realistic time-series data and practical benchmarks. This work introduces the first industrial-scale dataset that integrates maintenance events and operational mode transitions, enabling a systematic evaluation of steady-state and transient data-driven models alongside Bayesian filters. Furthermore, it pioneers the application of self-supervised learning to extract latent health representations without access to ground-truth health labels, thereby establishing a practical performance lower bound for this task. Experimental results demonstrate that conventional Bayesian filters remain strong baselines, while self-supervised approaches effectively uncover the intrinsic complexity of health state estimation. The study releases its dataset, code, and benchmark to advance interpretable health monitoring research.

Abstract

Estimating the health state of turbofan engines is a challenging ill-posed inverse problem, hindered by sparse sensing and complex nonlinear thermodynamics. Research in this area remains fragmented, with comparisons limited by the use of unrealistic datasets and insufficient exploration of the exploitation of temporal information. This work investigates how to recover component-level health indicators from operational sensor data under realistic degradation and maintenance patterns. To support this study, we introduce a new dataset that incorporates industry-oriented complexities such as maintenance events and usage changes. Using this dataset, we establish an initial benchmark that compares steady-state and nonstationary data-driven models, and Bayesian filters, classic families of methods used to solve this problem. In addition to this benchmark, we introduce self-supervised learning (SSL) approaches that learn latent representations without access to true health labels, a scenario reflective of real-world operational constraints. By comparing the downstream estimation performance of these unsupervised representations against the direct prediction baselines, we establish a practical lower bound on the difficulty of solving this inverse problem. Our results reveal that traditional filters remain strong baselines, while SSL methods reveal the intrinsic complexity of health estimation and highlight the need for more advanced and interpretable inference strategies. For reproducibility, both the generated dataset and the implementation used in this work are made accessible.

Problem

• turbofan health estimation
• inverse problem
• sparse sensing
• nonlinear thermodynamics
• temporal information

Innovation

• self-supervised learning
• inverse problem
• turbofan health estimation
• Bayesian filters
• realistic dataset