A comparative analysis on HHT and ANFIS in nonlinear noise cancellation

The applications of HHT and ANFIS were studied in nonlinear noise cancellation. Examples were given and compared. The results show that HHT can effectively cancel the high frequency noise when the noise has not too much amplitude overlapped with the original signal, while ANFIS still can do a good filtering job on the too much noise-overlapped signal when HHT fails. The main reasons of the phenomena were discussed. ANFIS benefits from its adaptive neuro-fuzzy inference ability, which is different from the frequency analysis method such as HHT. At last, some useful conclusions were given.