Skip to Main Content
More and more state variables in the case, the filter state variables, and identify operating modes means that the state collect and spend a lot of computing time, which lost control of real-time. And contains a lot of noise in electronic devices such as exceptions, these factors will influence the support vector machine to establish the optimal classification surface. High-frequency signal conversion equipment needs in the shortest possible time, alarm, address this requirement, we use a Huffman coding on the control signal compression, Then use a Kernel density estimation method, a structural form of fuzzy membership function, the membership function applied to the fuzzy support vector machines for fault diagnosis, This method can eliminate the characteristics of the impact of noise and outliers, through training support vector machines, we can get fault diagnosis model to realize the failure of electronic equipment, diagnostic classification. The method is applied to high-frequency signal conversion equipment for fault diagnosis, the results show that the compression algorithm used to retain equipment operation to the maximum extent, while greatly reducing the information processing time, Fuzzy support vector functions highlight the different characteristics of fault and correctly diagnose the fault type and effective, this method of fault diagnosis of electronic devices to provide a new way.