Diagnosis methodology for identifying gearbox wear based on statistical time feature reduction

Abstract

Strategies for condition monitoring are relevant to improve the operation safety and to ensure the efficiency of all the equipment used in industrial applications. The feature selection and feature extraction are suitable processing stages considered in many condition monitoring schemes to obtain high performance. Aiming to address this issue, this work proposes a new diagnosis methodology based on a multi-stage feature reduction approach for identifying different levels of uniform wear in a gearbox. The proposed multi-stage feature reduction approach involves a feature selection and a feature extraction ensuring the proper application of a high-performance signal processing over a set of acquired measurements of vibration. The methodology is performed successively; first, the acquired vibration signals are characterized by calculating a set of statistical time-based features. Second, a feature selection is done by performing an analysis of the Fisher score. Third, a feature extraction is realized by means of the Linear Discriminant Analysis technique. Finally, fourth, the diagnosis of the considered faults is done by means of a Fuzzy-based classifier. The effectiveness and performance of the proposed diagnosis methodology is evaluated by considering a complete dataset of experimental test, making the proposed methodology suitable to be applied in industrial applications with power transmission systems.

Novel Methodology for Condition Monitoring of Gear Wear Using Supervised Learning and Infrared Thermography

“…However, most of the reported methodologies are limited to the diagnosis of gearbox faults as a single gear tooth damage, whereas most analyzed conditions are related to the presence of irregularities in the teeth of the gears such as tooth breakage; chipping and cracks in the root; and chipping, pitting and damage to the surface of the tooth. [7,[19][20][21]. In fact, from an industrial viewpoint, these faulty conditions may be considered as a critical condition in gearbox transmission systems since their occurrence may produce critical damage to the entire transmission system, leading to the machine breakdown.…”

“…On the other hand, a preview research work that assessed uniform wear in gears has also been performed by our research group [7]; in this study, the assessment of the uniform wear in gears was carried out by means of vibration signals. The novelty of this work includes the consideration of feature selection and feature reduction stages, and the final diagnosis outcome was performed by means of a fuzzy-based classifier, resulting in an average of 97.27% for the global classification ratio.…”

“…Various faults may occur within the gearbox transmission system. In the literature, it has been reported that the occurrence of faults in the gears represents 80% in transmission machinery systems and 10% in rotating machinery systems [7], resulting from several factors ranging from inadequate lubrication, fluctuating loads, decoupling, misalignment, poor cooling, and gear design, among others [8]. In the literature, different methodologies for monitoring and detecting gearbox faults have been reported based on different sensing techniques and methods for analyzing and classifying data.…”

“…To detect these problems, adaptive wavelet filters are effectively applied to vibration signals acquired from the monitoring of the gearbox [20]. In addition, diagnosis methodologies based on vibration signals are proposed for the identification of wear in gears [7], with MCSA and load torque signature analysis (LTSA) being applied to detect faults such as misalignment, wear and mass imbalances of gears [21]. Nevertheless, few reported papers analyze and study gradual and uniform wear in the gears of the transmission system.…”

Short-Circuited Turn Fault Diagnosis in Transformers by Using Vibration Signals, Statistical Time Features, and Support Vector Machines on FPGA

“…For feature extraction, the classical methods include linear discriminant analysis (LDA), principal component analysis (PCA), and multidimensional scaling (MDS) [ 39 ]. On the other hand, a widely used feature reduction method is the Fisher score, which removes redundant features [ 40 , 41 ].…”