Soft Fault Diagnosis Technology of Filter Circuit Based on Heterogeneous Ensemble Learning

Recently, with an aim to improve the fault diagnosis accuracy of filter circuit in electronic equipment, ensemble learning method, a branch of machine learning has been emerging as a promising tool for applications with electronic area. To uncover its potentials in filter circuit, a soft fault diagnosis method based on heterogeneous ensemble learning is illustrated in this paper. The proposed method does not need to analysis the underlying transfer function of the circuit. Alternatively, it explored an effective ensemble model for fault diagnosis with three weak learners: linear regression, support vector machine (SVM) and decision tree, which establish the mapping function between fault features and output voltage. Then the fault classification also can be achieved by the proposed model. Finally, the experimental results with Sallen-Key filter circuit demonstrate that the proposed method outperforms several state-of-the-art methods.