Multiple Classifiers Combination Model for Fault Diagnosis Using Within-class Decision Support

In order to improve the reliability of fault detection and diagnosis for dynamic system, it is important to make full use of the information from different component of system. Multiple classifiers fusion is a technique that combines the decisions of different classifiers as to reduce the variance of estimation errors and improve the overall classification accuracy. This paper proposes a novel multiple classifiers fusion using within-class decision support for fault diagnosis. The new approach considers the fault diagnosis problem in time series. Then, one-step time series within-class decision support value and synchronization within-class decision support value are calculated to get association probability of each classifier in the same class recognition. Finally, calculate the fusion posterior probability outputs and normalize them for final decision. Experimental results demonstrate that the method is able to achieve a preferable solution, which has a better classification performance compared to single classifier.