Application of improved FastICA algorithm in compound fault diagnosis of vibrating scree

In view of problem that common fault characteristic extraction methods could not extract compound fault characteristic completely and effectively, an improved fast independent component analysis(FastICA) algorithm was proposed. The improved FastICA algorithm adaptively selects different nonlinear function to take progressive analysis, so data characteristic extracted by the improved FastICA algorithm is more accurate than that extracted by FastICA algorithm. The improved FastICA algorithm was applied to compound fault diagnosis of vibrating screen. The simulation and actual test results show that the improved FastICA algorithm can extract different fault characteristics effectively with higher separation accuracy.