Detection of bearing failure in rotating machine using adaptive neuro-fuzzy inference system

This paper proposes a novel approach for bearing health evaluation using Lempel-Ziv Complexity and time domain statistical parameters in conjunction with ANFIS. Compared to conventional techniques the presented approach works well for a non linear physical system and is thus suited for condition monitoring of machine system under varying operating and loading conditions. The performance of this technique is investigated through experimental study of realistic vibration signals. The results demonstrate that complexity analysis and time domain parameters in conjunction with ANFIS provide an effective measure forbearing health evaluation.