| In this paper, the general situation of the fault diagnosis technology and the intelligent fault diagnosis system in the recent years is introduced, and the various existing fault diagnosis methods are analyzed and evaluated. In the end, the rationale for diagnosis is introduced and the existing problems of the fault diagnosis system are proposed. The focus of the study is on the reduction and improvement of algorithm on the basis of the discernibility matrix attribute, on the combination of this improved algorithm in the rough set theory with the neural network theory, as well as on their application in the fault diagnosis of the rotating machinery and bearings.With fault diagnosis of rotating machinery and bearings as example, the improvement of algorithm on the basis of the discernibility matrix attribute is applied to reduction of incomplete decision system to find necessary conditions for diagnosis. Based on the optimal decision system obtained from the reductions, Elman and BP neural network are designed for fault identification. The application of the reduced decision system to the neural fault classifier indicated that the improvement of algorithm based reduction reduces the dimension of input to neural network, and raises the efficiency of training and computation.The main works are listed as follows: 1. Greatly reduced the amount of computation and the risk of errors through the improvement of algorithm on the basis of the discernibility matrix attribute.2. Mainly discussed the differences between the compatible and the incompatible decision-making table in calculating the discernibility matrix. Explored the possible errors in the calculation of matrix with the incompatible decision-making table and figured out the crux of the problem as well as the possible improvement. Proposed the improving methods of algorithm on the basis of the discernibility matrix attribute in the incompatible decision-making table.3. Combined the rough set theory with the Elman network and applied the combination in the fault diagnosis of rotating machinery. The Matlab simulation results were given and the superiority of this improved algorithm was discussed.4. Studied the bearing fault diagnosis, processed the data with the improvement of algorithm on the basis of the discernibility matrix attribute, and applied this improved algorithm in fault diagnosis through the BP network. Matlab simulation results showed that the fault diagnosis method stated in the paper was absolutely right. Through a comparison study with other methods in bearing fault diagnosis, the paper pointed out that this method can decrease the computation time and increase the diagnosis correctness. |
