Stalled pressure data in compressor was de-noised by wavelet analysis and then resolved with multi-resolution analysis.The pressure signals in related frequency band were reconstructed by the wavelet coefficient in dN level.Rotating stall was considered to be examined out when the amplitude of the reconstructed signals was somewhere beyond the given threshold value.The results show that it is effective to identify the peak signal and noise by wavelet decomposition.Different threshold values are given with different sampling frequencies.By monitoring the dN level of the reconstructed signals

rotating stall can be detected.