Experimental Investigation and Fault Diagnosis for Buckled Wet Clutch Based on Multi-Speed Hilbert Spectrum Entropy

Xue, Jiaqi; Ma, Bin; Chen, Man; Zhang, Qianqian; Zheng, Liangjie

doi:10.3390/e23121704

Cited by 11 publications

(6 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using both low-frequency and high-frequency components that may be indicative of various fault types, wavelet analysis proves to be a multi-resolution method. The complexity and irrationality of the time-frequency distribution of the signal are measured by the Hilbert spectrum entropy, which sheds light on how the clutch system vibrates [12,13].…”

Section: Anomaly Detectionmentioning

confidence: 99%

Deep learning based fault detection of automobile dry clutch system using spectrogram plots

Sai S,

Sridharan,

Chakrapani

et al. 2024

Eng. Res. Express

View full text Add to dashboard Cite

In this research study, a novel approach that leverages the power of transfer learning (a famous deep learning technique) is proposed to diagnose multiple types of clutch faults including, worn release fingers, fractured pressure plates, deteriorated pressure plates, loss of friction material and distorted tangential strips using spectrogram plots. To train and validate the diagnostic system, vibration readings were taken from a specially designed test rig with the help of piezoelectric accelerometer while the clutch system was operated under different load conditions of 0 (no load), 5 and 10 kg This procedure of data collection was then repeated to acquire the vibration data for all of the fault conditions by replacing the good with fault components individually. These vibration signals were further processed and transformed into spectrogram plots that serves as the input data for the deep learning models considered. Fine-tuning techniques were applied on pretrained networks to maximise the prediction accuracy of the models to effectively determine and diagnose faults in the clutch system. For this study 12 pre-trained networks were chosen namely, Xception, InceptionResNet, DenseNet, AlexNet, VGG16, GoogLeNet, VGG19, ResNet101, ResNet50, InceptionV3, MobileNetV2 and ShuffleNet. To optimize the performance of deep learning models, a systematic adjustment of hyperparameters such as the train-test split ratio, learning rate, optimizer and batch size for each network model was carried out. Through careful experimentation and analysis, significant improvements in fault classification accuracy were achieved thereby enhancing the reliability and effectiveness of the diagnostic system. From the results it was noted that 100% classification accuracy was displayed by AlexNet (for the no load condition and the 10kg load condition) and GoogLeNet (for 5 kg load condition) with extremely low computation times.

show abstract

Section: Anomaly Detectionmentioning

confidence: 99%

Deep learning based fault detection of automobile dry clutch system using spectrogram plots

Sai S,

Sridharan,

Chakrapani

et al. 2024

Eng. Res. Express

View full text Add to dashboard Cite

show abstract

“…. x = Ax + Bu+D dis y = Cx + Du (7) where x denotes the state vector, y denotes the output, and u denotes the input. The continuous state space equation needs to be discretized.…”

Section: Mpcmentioning

confidence: 99%

“…In addition, HSCS is a nonlinear system. The key to the rotational speed control loop is to overcome serious nonlinear problems such as large hysteresis, large dead zone, and time-varying parameters of the hydro-viscous clutch [7][8][9]. It is necessary to dynamically change the control parameters according to the changes in the model and working conditions [10,11].…”

Section: Introductionmentioning

confidence: 99%

A Novel Double Closed Loop Control of Temperature and Rotational Speed for Integrated Multi-Parameter Hydro-Viscous Speed Control System (HSCS)

Zhao,

Wang,

Wang

et al. 2024

Machines

View full text Add to dashboard Cite

Hydro-viscous clutch has already become an inevitable choice for special vehicle transmission in the present and future. As a nonlinear system with a large hysteresis loop, its speed control performance is affected by input rotational speed, lubricating oil temperature, lubrication pressure, and other factors. The traditional control method cannot adjust the temperature and rotational speed, which will lead to problems of narrow speed range, poor rotational speed stability, and large dynamic load impact. In order to solve the above problems, this paper studies the control method of an integrated multi-parameter hydro-viscous speed control system (HSCS) in a controlled environment. Through the mechanism analysis of the law of HSCS, the influence law of speed and temperature during the system operation is found. The temperature closed loop based on model predictive control (MPC) is introduced to control the rotational speed, and then the traditional PID control results are compensated according to the speed closed loop. Next, a novel double closed loop control method of temperature and rotational speed for HSCS is formed. Finally, the simulating verification is carried out. Compared with the traditional control method, the design method in this paper can adjust the control parameters according to the temperature of the lubricating oil and the input rotational speed and effectively expand the domain of HSCS and the speed control stability. The effective transmission ratio is extended to 0.2~0.8, and the hydro-viscous torque and speed fluctuation under the engine rotational speed fluctuation are reduced by more than 30%. The novel control method of HSCS designed in this paper can effectively improve the influence of input rotational speed and lubricating oil temperature on the speed control performance of HSCS and can be widely used in nonlinear HSCS such as hydro-viscous clutch.

show abstract

“…The results of the computerized wear prediction model were tested by sensitivity analysis and model validation using Monte Carlo method. Xue (2021) carried out bench test on the clutch to obtain vibration signals of each type at different speeds. By comparing the accuracy of different classifiers with and without entropy, the buckling state of friction parts can be identified, and then the clutch state can be judged.…”

Section: Introductionmentioning

confidence: 99%

Residual life prediction of wet clutch based on binary Wiener process

Liu

Zhang

Cui

et al. 2023

ILT

View full text Add to dashboard Cite

Purpose In armored vehicles integrated transmissions, residual life prediction based on oil spectrum data is crucial for condition monitoring and reliability assessment. This paper aims to use the advantages of real-time and accurate prediction of binary Wiener process, the residual life prediction of clutch is studied. Design/methodology/approach First, combined with the wet clutch life test, the indicator elements Cu and Pb and the failure threshold of the residual life prediction of the clutch are extracted through the oil replacement correction of the spectral data of the whole life cycle; second, the correlation characteristics of indicating elements are analyzed by MATLAB Copula function, then the correlation function of residual life will be derived; third, according to the inverse Gaussian principle, the performance degradation mathematical models of the unary and binary Wiener processes of the above two indicator elements are established; finally, the maximum likelihood estimation method is used to estimate the parameters, and the monadic and binary performance degradation mathematical models are used to predict the residual life of the tested clutch. Findings By comparing the prediction results with the test results, with the passage of time, 81.25% of the predicted value error of the residual life prediction method based on the binary Wiener process is controlled within 20%, while 56.25% of the predicted value error of the residual life prediction method based on the unitary Wiener process is controlled within 20%. At the same time, the prediction accuracy of the binary prediction model is 2%–16.7% higher than that of the unitary prediction model. Originality/value This paper studies the residual life prediction theory of wet clutch, which can develop the theory and method of comprehensive transmission health monitoring, and provide theoretical and technical support for the construction of a reliable health management system for high-speed tracked vehicles.

show abstract

Experimental Investigation and Fault Diagnosis for Buckled Wet Clutch Based on Multi-Speed Hilbert Spectrum Entropy

Cited by 11 publications

References 32 publications

Deep learning based fault detection of automobile dry clutch system using spectrogram plots

Deep learning based fault detection of automobile dry clutch system using spectrogram plots

A Novel Double Closed Loop Control of Temperature and Rotational Speed for Integrated Multi-Parameter Hydro-Viscous Speed Control System (HSCS)

Residual life prediction of wet clutch based on binary Wiener process

Contact Info

Product

Resources

About