Intelligent chatter detection based on wavelet packet node energy and LSSVM-RFE

Chatter is an unstable phenomenon in machining processes which brings many detrimental effects to cutting tool and workpiece. This paper proposes an intelligent chatter detection method which is based on wavelet packet node energy (WPNE), least squares support vector machine based recursive feature elimination (LSSVM-RFE) and least squares support vector machine (LSSVM). The method consists of three steps. First, feature vectors, with high resolution of energy distribution, are constructed by the sixth level WPNE instead of the third level WPNE to acquire more information of machining processes. Second, a feature selection method, LSSVM-RFE, is presented to select the most outstanding features from the original feature set. Finally, the most outstanding features are fed into LSSVM to detect chatter. Turning experiments are conducted to verify the effectiveness and reliability of the proposed approach. Moreover, the effectiveness of the sixth level WPNE and LSSVM-RFE are also demonstrated by the experimental results.