Fault diagnosis method of rotating bearing based on improved ensemble empirical mode decomposition and deep belief network

Zhong, Cheng; Wang, Jie-Sheng; Sun, Weiyuan

doi:10.1088/1361-6501/ac6cc9

Cited by 22 publications

(11 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The parameter search range during the final optimisation process was determined based on this interval range. The search range for the penalty factor α was defined as [3000, 8000], whereas the range for the number of decomposition layers K was established as [3,15]. The number of searched individuals is 10, the maximum iteration number is 20, and the migration coefficient is 40.…”

Section: Data Preprocessingmentioning

confidence: 99%

“…The initial phase of the data-driven health condition evaluation method involves signal preprocessing. In contemporary research, prevalent signal preprocessing techniques encompass the wavelet packet transform (WPT), empirical mode decomposition (EMD) and ensemble EMD (EEMD) [15]. Zhou et al proposed an adaptive wavelet denoising method to decompose the vibration signals of rotating machinery [16].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optimised LightGBM-based health condition evaluation method for the functional components in CNC machine tools under strong noise background

Jia,

Jialong,

Wanghao

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

The accurate health condition evaluation of the functional components in computer numerical control (CNC) machine tools is an important prerequisite for predictive maintenance and fault warning. The vibration signals of the functional components in CNC machine tools often contain substantial noise, impeding the extraction of relevant health condition information from the vibration signals. This work presents an approach that leverages the variational mode decomposition (VMD) enhanced by the Artificial Hummingbird Algorithm (AHA) alongside the Light Gradient Boosting Machine (LightGBM) optimised through particle swarm optimisation (PSO) to evaluate the health condition of the functional components in CNC machine tools amidst pervasive noise. Initially, the AHA optimised the penalty factor (α) and the decomposition layer (K) within the VMD. This optimised VMD was subsequently applied to denoise the original vibration signals. After this denoising process, PSO was employed to optimise the learning rate and maximum tree depth within LightGBM. Health condition evaluation experiments were executed on the feed system and spindle of the CNC machine tool to validate the proposed methodology. Comparative analysis indicates that the proposed method attains paramount accuracy and computational efficiency, which are crucial for accurately evaluating the health condition of the functional components in CNC machine tools.

show abstract

Section: Data Preprocessingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Optimised LightGBM-based health condition evaluation method for the functional components in CNC machine tools under strong noise background

Jia,

Jialong,

Wanghao

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…As classical signal processing methods, EEMD and CEEMDAN are often used by researchers to compare with the proposed methods. 48–50 To verify the effectiveness of ICEEMDAN in fault diagnosis, the original vibration signal and the feature vectors extracted by EEMD, CEEMDAN, and ICEEMDAN are used as the input of Ghost-IRCNN model, and the accuracies are 95.33%, 97.75%, 98.27%, and 99.74% respectively. The feature vectors extracted by ICEEMDAN has the highest diagnostic accuracy in Ghost-IRCNN model, which is 4.44%, 1.99%, and 1.47% higher than the other three methods.…”

Section: Experimental Analysismentioning

confidence: 99%

Intelligent diagnosis method of bearing fault based on ICEEMDAN and Ghost-IRCNN

Zou

Jin

et al. 2022

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

As a critical component of rotating machinery, the health status of bearings is of great significance to ensure the safety of machine operation. Since the vibration signals are interfered by noise during the machine operation, the traditional bearing fault diagnosis model has insufficient feature extraction ability, low diagnostic accuracy, and slow convergence speed. An intelligent bearing fault diagnosis method is proposed based on improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) and inverted residual convolution neural network integrated with Ghost module (Ghost-IRCNN). Firstly, ICEEMDAN is used to decompose the signals, and the kurtosis-correlation coefficient-margin factor index is proposed to filter and reconstruct the important modal components. Then, the multi-scale sample entropy is used to construct the feature vectors. Finally, the Ghost module, convolution block attention module, and inverted residual bottleneck are combined to build Ghost-IRCNN to realize bearing fault classification and diagnosis. The experimental results show that the proposed method can extract bearing fault features effectively, and the diagnosis accuracy can reach more than 98.80% under complex conditions.

show abstract

“…In the signal processing field, most of the traditional weak signal detection methods attempt to suppress or eliminate the complex and changeable noise embedded in the received signals to extract the weak signal characteristics; these methods primarily include wavelet transform [1,2], empirical mode decomposition [3][4][5], independent component analysis [6,7] and other filters. However, while reducing the noise, these approaches also weaken the weak signal, which makes the detection performance unsatisfactory.…”

Section: Introductionmentioning

confidence: 99%

Adaptive progressive learning stochastic resonance for weak signal detection

zong¹,

Men²,

An³

et al. 2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Stochastic resonance (SR) can enhance signals by using noise. This has attracted more attention in the field of weak signal detection. In practical applications, owing to the non-adjustability of noisy signals, SR is required to adjust the system parameters adaptively to satisfy the conditions of the SR phenomenon. In this paper, an adaptive progressive learning stochastic resonance (APLSR) method is proposed to improve the detection ability of weak signal, and the SR phenomenon is quantitatively defined. A theoretical learning framework is established with an improved reinforcement learning model by mapping the nonlinear system parameter space to a progressive learning set. By selecting a proper learning layer within a determined constraint range, the matching system parameters can be quickly and accurately searched to generate a desired optimal output. Numerical simulation results show that the signal energy and the output-SNR can be enhanced significantly, which reflects an excellent weak signal detection performance especially for low signal-to-noise ratio (SNR) conditions. Finally, a diagnosis on the outer race fault signals of rolling bearing confirms that the proposed method can effectively detect fault characteristics.

show abstract

Fault diagnosis method of rotating bearing based on improved ensemble empirical mode decomposition and deep belief network

Cited by 22 publications

References 60 publications

Optimised LightGBM-based health condition evaluation method for the functional components in CNC machine tools under strong noise background

Optimised LightGBM-based health condition evaluation method for the functional components in CNC machine tools under strong noise background

Intelligent diagnosis method of bearing fault based on ICEEMDAN and Ghost-IRCNN

Adaptive progressive learning stochastic resonance for weak signal detection

Contact Info

Product

Resources

About