Combine Assembly Fault Diagnosis Based on Optimized Multi-scale Reverse Discrete Entropy

An optimized multi-scale reverse discrete entropy (RDE, OMRDE) method for feature extraction is proposed to address the lack of effective feature extraction and detection methods for combining harvester assembly fault inspection. This method is used to extract vibration signal features from the harvester. A fault diagnostic method is designed to verify the efficiency of the associated methods. First, a comparative study of RDE, multi-scale inverse RDE (MRDE), and OMRDE was performed using simulated signals to verify the effectiveness of OMRDE. Second, the FSTPSO–VMD method was used to decompose the vibration signal of the combine harvester assembly fault, and the OMRDE, MRDE, and fuzzy entropy were compared and analyzed. The actual feature extraction effect of the three entropy functions reached the highest classification accuracy (88.5%) after using OMRDE to extract features. Finally, a fusion feature set is constructed to further improve the classification accuracy, and the LSSVM classifier is further optimized through FSTPSO. Analytical results show that the FSTPSO–LSSVM classifier constructed in this work adopts the fused feature with an accuracy of 93%, which is better than other common methods and verifies the validity of the fault diagnostic model. Therefore, the performance of the OMRDE proposed in this work is better than those of MRDE and MRDE. The proposed fault diagnostic model can realize accurate classification of the combine harvester assembly fault detection.

A Simultaneous Fault Diagnosis Method Based on Cohesion Evaluation and Improved BP-MLL for Rotating Machinery

In this paper, an improved simultaneous fault diagnostic algorithm with cohesion-based feature selection and improved backpropagation multilabel learning (BP-MLL) classification is proposed to localize and diagnose different simultaneous faults on gearbox and bearings in rotating machinery. Cohesion evaluation algorithm selects high sensitivity feature parameters from time and frequency domain in high-dimensional vectors to construct low-dimensional feature vectors. The BP-MLL neural network is utilized for fault diagnosis by classifying the feature vectors. An effective global error function is proposed in BP-MLL neural network by modifying distance function to improve both generalization ability and fault diagnostic ability of full-labeled and nonlabeled situations. To demonstrate the effectiveness of the proposed method, simultaneous fault diagnosis experiments are conducted via wind turbine drivetrain diagnostics simulator (WTDDS). The experiment results show that the proposed method has better overall performance compared with conventional BP-MLL algorithm and some other learning algorithms.