Xiaofeng Liu

Gas path fault diagnosis in high-performance gas turbines is often challenged by insufficient sensor availability, leading to underdetermined problems where fault isolation is easily confused with degradation in other components. In such cases, isolation results become unreliable when the number of sensors is smaller than the number of estimated health parameters, and multiple feasible solutions arise. To alleviate this issue, this paper introduces an isolability indicator for gas path faults under simultaneous unknown degradations, along with a sensor selection algorithm that maximizes isolability. The isolability of a component fault is defined as the probability of distinguishing it from faults in other components, and a sensor configuration performance metric is constructed from the isolabilities of all faults. Since calculating isolability through Monte Carlo (MC) simulations imposes a heavy computational burden, an approximation method based on clustering to reduce MC simulations is developed. A simplified indicator is designed to infer the change of isolability. And these simplified indicators can be classified by the geometric features of the fault symptoms sets. Building on this, an algorithm is proposed to determine the necessary MC simulations, further estimating the remaining isolability and reducing computational cost. The approach is validated on a high-bypass ratio turbofan engine model through digital simulations. Results demonstrate that the proposed indicator effectively characterizes component fault isolability, and that appropriate sensor selection enhances fault isolability in the presence of unknown degradation. Moreover, the algorithm successfully identifies optimal or near-optimal sensor configurations across different scenarios.

According to the multi-mode characteristics, aeroengines are usually modeled as switched systems to better reflect the working dynamics, and switched controllers are designed to fulfill different requirements for each subsystem, which may encounter the control bumps at controller-switching instants. To solve this issue, a class of mode-recovered bumpless transfer control is proposed for aero-engines, which not only can eliminate the control bumps, but also guarantees the recoverability of control signal for each subsystem. First, the mode-recovered compensator is newly proposed, which reshapes the control signal into a continuous and bumpless form. By using the fixed-time analysis, the bumpless control signal is ensured to recover its original value with a uniformly bounded settling time, which is no more than each mode-running interval. Subsequently, under the hybrid dwell-time switching regulation, the closed-loop system is addressed to be input-to-state stable. Furthermore, the proposed method is implemented to both the state feedback control and the output feedback control with solvable conditions to optimize the controller gains. Finally, applications to the turbofan aeroengine and the variable cycle engine are provided, which further reveal the effectiveness.

Turbofan engines are known as the heart of the aircraft. The turbofan’s health state determines the aircraft’s operational status. Therefore, the equipment monitoring and maintenance of the engine is an important part of ensuring the healthy and stable operation of the aircraft, and it is vital to monitor the remaining useful life (RUL) of the engine. The monitored data of turbofan engines have high dimensions and a long time span, which cause difficulties in predicting the remaining useful life of the engine. This paper proposes a residual life prediction model based on Autoencoder and a Temporal Convolutional Network (TCN). Among them, Autoencoder is used to reduce the dimension of the data and extract features from the engine monitoring data. The TCN network is trained on the obtained low-dimensional data to predict the remaining useful life. The model mentioned in this article is verified on the NASA public data set (C-MAPSS) and compared with common machine learning methods and other deep neural networks. The SAE-TCN model achieved better scores on the FD001 independent testing data set with an RMSE of 18.01 and a score of 161. The average relative error of the model relative to other common learning models is 0.9499 in RMSE and 0.2656 in Scoring Function. The experimental results show that the model proposed in this paper performs the best in the evaluation, and this conclusion has important implications for engine health.