Abstract
Background: Vibration signature acquired from a gear mesh can be used to identify the defect present in a gear mesh hence can be used to diagnose the condition of a gear mesh. But the signal acquired from the subject may not be noise free and may be non stationary.
Methods: Before going for the analysis of the acquired signal a preprocessing on the acquired signal is required to make it noise free. In the present work in first phase, the acquired vibration signal is filtered to reduce the noise and to improve the SNR (signal to noise ratio). The filtering is done by an Adaptive Noise Cancellation (ANC) technique. A modified Leaky Least Mean Square (LLMS) based adaptive algorithm along with a digital filter is used to achieve the ANC. The signal acquired from a healthy gear is used as the reference signal for the adaptive filter based de-noising process. In the second phase of the present work Adaptive Wavelet Transform (AWT) is used to detect the fault by extracting the features from the filtered vibration signal. From the signal pattern the adaptive wavelet is designed. The adaptive wavelet scalogram is compared with the standard wavelet scalogram.
Results: The result shows that the adaptive wavelet scalogram is better in analyzing the vibration signal. In this work a gear drive experimental set-up is made. Two different types of defective gears are used for the experiment. In type-1 defective gear one tooth is broken and in type-2 defective gear two teeth are broken. Initially, the vibration signal is acquired from a healthy gear which is used as the reference signal. Then the vibration signal from type-1 defective gear and type-2 defective gear is acquired and processed for the analysis and to identify the defects.
Conclusion: The present work shows that with the application of modified-LLMS algorithm and AWT the proposed technique of signal processing is more suitable for the fault identification and hence for the condition monitoring of the gear.
Keywords: Adaptive wavelet, Adaptive Noise Cancellation (ANC), gear fault, modified-LLMS, adaptive algorithm, leaky least mean square.
Graphical Abstract
Call for Papers in Thematic Issues
Advancements in Brain Tumor Detection and Treatment using Machine Learning, Deep Learning, and Blockchain Technology
Brain tumors are among the most challenging diseases to diagnose and treat, requiring specialized expertise and advanced technology for accurate detection and effective treatment. In recent years, machine learning and deep learning algorithms have shown promise in improving the accuracy and efficiency of brain tumor detection through medical imaging analysis. ...read more
Advances of Biometric Data Analysis to Boost Intelligent Applications
With the rapid development of artificial intelligence (AI) technology, more and more intelligent devices and applications are appearing in our daily lives, such as smart home, smart agriculture, health diagnosis, educational support, and environmental monitoring. Interaction with these devices or applications has become a pressing issue that can greatly improve ...read more
Attainment of SDGs through the Advancement in Solar Energy Systems
With less than a decade until we reach 2030, it is crucial to address the deep inequalities affecting not only our health but also our quality of life, and the economy of countries worldwide. Few of the UN's Sustainable Development Goals (SDGs) can be directly and indirectly achieved through the ...read more
Emerging Intelligent Computing Techniques and their Applications Using Machine Learning
This special issue will aim to discuss computational intelligence approaches, initiatives and applications in engineering and science fields (including Machine Intelligence, Mining Engineering, Modeling and Simulation, Computer, Communication, Networking and Information Engineering, Systems Engineering, Innovative Computing Systems, Adaptive Technologies for Sustainable Growth, and Theoretical and Applied Sciences). This collection should ...read more
Related Journals
Recent Patents on Engineering
18
4