Power-fold actuator noise diagnostics using statistical envelope analysis

A good correlation between subjective evaluation and objective test is a key for any machine condition monitoring and fault diagnosis process. This paper proposes a statistical envelope analysis approach for product quality control where Fourier spectrum is selected as diagnosis analysis characteristics. The acceptance amplitude envelope is statistically calculated from the measured random spectrum data of a large number of subjectively good samples. A product is then assessed by comparison of its measured spectrum data to the acceptance envelope. If the spectrum amplitude stays within the envelope, the product is accepted as good one, otherwise it is rejected. The statistical envelope analysis approach has been validated by diagnosis test examples of power fold (PF) actuators. Based on the correlation of objective tests and subjective evaluations of given samples, the diagnosis methodology illustrated in this paper is applicable to other products for noise diagnosis and quality control.