Fault Diagnosis of Loader Gearbox Based on an ICA and SVM Algorithm

When a part of the loader's gearbox fails, this can lead to equipment failure due to the complex internal structure and the interrelationship between the parts. Therefore, it is imperative to research an efficient strategy for transmission fault diagnosis. In this study, the non-contact characteristics of noise diagnosis using sound intensity probes were used to collect noise signals generated under gear breaking conditions. The independent component analysis (ICA) technique was applied for feature extraction from the original data and to reduce the correlation between the signals. The correlation coefficient between the independent components and the source data was used as the input parameters of the support vector machine (SVM) classifier. The separation of the independent components was achieved by MATLAB simulation. The misdiagnosis rate was 5% for 40 sets of test data. A 13-point test platform for noise testing of the loader gearbox was built according to Chinese national standards. Source signals under the normal and fault conditions were analyzed separately by ICA and SVM algorithms. In this case, the misdiagnosis rate was 7.5% for the 40 sets of experimental test data. This proved that the proposed method could effectively realize fault classification and recognition

Algorithm · Blind signal separation · Classifier (UML · Fault (geology · Feature extraction · Independent component analysis · Loader · MATLAB · Noise (video · Pattern recognition (psychology · Support vector machine · Test data · Advanced Algorithms and Applications · Advanced Computational Techniques and Applications · Computer Science · Machine Fault Diagnosis Techniques · Artificial Intelligence

• A Hybrid PSO–SVM Model Based on Safety Risk Prediction for the Design Process in Metro Station Construction

  Open Access•Ping Liu, Mengchu Xie et al.•International Journal of…•2020

  Add to list