Paper Title
An Application of Lathe Machine Chatter Analysis based on Chua’s Circuit and Fractional-Order Lorenz Master-slave Hybrid Chaotic System

Abstract
Lathe machine has always played an important role in the precision machining and manufacturing industry. To meet the increasing demand for higher machining quality, the improvement on smart machine tools and development of intelligent peripherals have become a trend. However, tool wear and chatter are the most crucial key factors, among others, that affect the machining quality. They also affect the appearances and tolerance of finished products and shorten tool life, leading to cost increase. This study is focused on the subject of the chattering of machines. Traditionally, chatter identification is based on operator’s experience instead of quantized data. The disadvantages include the lack of unified standard and high misjudgment rate. A great number of further studies have been carried out on the two phenomena above in recent years. Most of the identification methods are based on finding characteristics of the phenomena in time domain and frequency domain. Frequency-domain methods, although more popular, are inappropriate for real-time analysis due to the drawbacks such as high time complexity, unfitness for straightforward chatter identification, low computational efficiency caused by the need for high-dimensional data. This study is focused on the phenomenon of chattering of lathe machines. An analysis method that combines Chua’s circuit, fractional-order Lorenz master-slave hybrid chaotic system and probabilistic neural network (PNN) is proposed. By comparing the dynamic errors generated by the chaotic systems of different orders, the fractional order of the chaotic system that produces the largest variation of characteristics is chosen. This approach results in a straightforward identification method based on images. Then the classification is carried out using PNN. Compared to frequency analysis method, the method proposed in this paper has advantages including lower-dimensional data and less computation needed, better efficiency, more straightforward and high-accuracy identification. The experiment results indicate that a 100% success rate of chatter identification can be reached. Index Terms- Fractional-order chaotic system, Chua's circuit, Lorenz chaotic system, Probabilistic neural network, Turning chatter vibration