The Application of LMS Filter Based on Initial Value Estimated by Fuzzy Inference for Active Vibration Control

The fuzzy system models with frequency of excitation signals for initial tap weight vector, which was used in the least mean square (LMS) adaptive filter, were built based on fuzzy inference approach. The model outputs were chosen as initial tap weight vector, which sped the convergence and improves the stability of the algorithm. The approach presented was tested by the experiment for the active vibration control of the single degree-of-freedom system. The results show the efficiency of the new approach.