26 October 2021 in IEEE Transactions on Instrumentation and Measurement
Using transfer learning (TL) for fault detection and diagnosis has been a hot topic in prognostic and health management (PHM) field. In this article, a systematic framework is established to solve the cross-domain fault diagnosis for rotary machines of the same type using entropy-based TL. The multiscale symbolic dynamic entropy (MSDE) is first proposed to extract features from vibration signals. Then, TL model is trained to get a mapping matrix, which can preserve the structure properties of prior distribution and minimize the discrepancies between different datasets. The mapped features are called multiscale transfer symbolic dynamic entropy (MTSDE). Finally, the support vector machine (SVM) classifier is utilized to accomplish the cross-domain intelligent fault-type recognition. To demonstrate the superiority of the proposed MTSDE method, the comparative experiments of nine methods are carried out under different datasets including gearboxes and bearings. Experimental results demonstrate that our proposed MTSDE method performs best in recognizing various fault types comparing with other nine methods. To the best knowledge of the authors, this is the first attempt of using entropy-based TL for cross-domain fault diagnosis of rotating machinery.
21 March 2023 in IEEE Sensors Journal
Due to coupling of multiple cylinders during operation and association of high environmental noise, the existing waveform analysis is inefficient to identify the working status of large multi-cylinder marine diesel engines (MCMDE). In order to solve this problem, a new framework named intrinsic multiscale dispersion entropy (IMDE) is proposed by calculating the multiscale dispersion entropy (MDE) of intrinsically reconstructed instantaneous angular speed (IAS) signal. Firstly, intrinsic characteristic scale decomposition (ICD) has been utilized to decompose an IAS signal into principal components. Then, appropriate components are selected for reconstruction of the intrinsically denoised IAS signal. Lastly, working state of the MCMDE is identified by quantifying the intrinsically reconstructed IAS signal with the help of multiscale dispersion entropy (MDE) leading to concept of IMDE. Simulation model corresponding to a v-16 cylinder engine and experimental data collected from a real life v-16-cylinder marine diesel engine are utilized for validation. Results show that the proposed IMDE can extract the effective fault features under different working conditions and has the highest classification accuracy in compare to other existing techniques: MDE, MSE and MFE.
1 January 2021 in Mechanical Systems and Signal Processing
16 June 2025 in IEEE Transactions on Instrumentation and Measurement
Achieving information fusion of multi-sensor data plays an important role in improving the performance of gearbox fault diagnosis. However, this fusion process is hindered by the heterogeneity problem caused by the different data dimensions of various sensors. In order to solve this problem, exploitation of the complementary nature of multi-source heterogeneous data to provide more accurate fault information is necessary. Thus, a multi-source heterogeneous information fusion method based graph convolutional network (MHIF-GCN) is proposed in this paper. In this framework, a convolutional autoencoder is used to extract deep features corresponding to different types of sensors as graph node features for solving data heterogeneity problem. Secondly, graph convolutional network (GCN) model based on K-nearest neighbor graph (KNNGraph) are introduced to establish the connection between different sensor data in the graph structure for realizing the feature-level fusion of sensor data and mining deeper fault data features. The results of two gearbox experiments validate the excellent fault diagnosis performance of the proposed MHIF-GCN. In Experiment I, the MHIF-GCN is able to accurately recognize six structural and nonstructural fault types. With the support of the complementary fusion mechanism, the proposed MHIF-GCN has the highest average diagnostic accuracy of 99.00% when compared with the other six methods. Even with a small number of training samples, MHIF-GCN still performs very favorably compared to other methods with an accuracy of 88.87%. In Experiment II, the MHIF-GCN has the highest diagnostic accuracy of 94.00% and the recall, precision, and f-score for each fault state remain above 85%, and the proposed MHIF-GCN maintains a stable diagnostic performance.
1 July 2021 in Ocean Engineering