Fault diagnosis method for machinery based on multi-source conflict information fusion

Abstract

Multi-source information fusion diagnosis is usually more reliable than the fault diagnosis with single-source employed. However, fusion resultsmaybe absurdity while fusing the highly conflicting information. To address this problem, the DS evidence theory is updated by weighting each evidence according to the corresponding contribution to the decision and a novel fault diagnosis method based on multi-source conflict information fusion is proposed. First, the basic probability assignment of evidence corresponding to the sensor information is given by introducing the feature parameters of electromyographic signals and using the BP neural network. Then, the importance of each evidence is determined by solving the difference degree and exclusion degree among the evidence, and the evidences are assigned weights according to the importance degree of each evidence in the fusion decision-making process. Next, weighted evidence are combined for making decision and further diagnosis after weighted averaging the evidences with different weights. In the end, the performance of the proposed method is assessed by the receiver operating characteristic (ROC). The experimental results show that the area under the ROC curves the proposed method are 0.3229, 0.0729, and 0.9271, higher than those of the traditional DS’s method, Murphy’s method, and Yager’s method, respectively, which proves that the proposed method has better diagnostic performance and reliability.

Designing and Manufacturing of Industrial Robots with Dual-Angle Sensors Taking into Account Vibration Signal Fusion

“…proposed to combine multivariate statistics and the artifcial intelligence algorithm to achieve the fusion of multisource vibration signals. Tey propose to combine multivariate statistical methods and fusion accuracy to derive the optimal assignment of weighted averaging [9]. Other scholars have reconstructed for vibration signal combustion information combined with multichannel vibration acceleration signal, and the signal with high correlation of combustion information is obtained in a new way, while the fusion of its vibration acceleration signal is achieved by weighting coefcients [10].…”

Designing an Improved VMD-CNNs-DS Algorithm for the Port Crane Bearing Fault Diagnosis

“…This study adopts Wei's proposed method that integrates the degree of difference and exclusion between evidence into the decision-making process, aiming to overcome the constraints of DS evidence theory [23]. Firstly, we calculate the degree of discrepancy and exclusion between each piece of evidence to determine their importance.…”

Application of multi-kernel relevance vector machine and data pre-processing by complementary ensemble empirical mode decomposition and mutual dimensionless in fault diagnosis

“…For the sample with the category label i, the expected estimated value Ỹin of y in is updated by equation ( 16). For samples of other categoriesb (∀b ̸ = i), the estimated value Ỹbn of y bn is updated by equation (17).…”

“…Therefore, research on efficient feature extraction methods is fundamental to realizing intelligent fault classification [15]. Many studies have analyzed the signal in the time and frequency domains, which retains not only the original characteristics of the signal but also interference in the signal [16,17]. Therefore, characteristic structures of time-domain signals, such as dimensionless indicators, have been constructed [18].…”