【摘要】 切削颤振是制约机械加工质量和效率的主要因素,传统的颤振监测方法在早期颤振的快速识别方面存在不足,提出一种利用在线监测车削过程中声压信号的小波包能量峭度指标来预警早期切削颤振。从时频域角度分析切削过程从稳定到非稳定状态的振动特点,采用小波包对采集的声压信号进行实时分解,挖掘声音信号中早期颤振孕育的微弱特征。计算每个频带节点的小波包能量,构造小波包能量峭度指标来反映早期颤振的孕育状态,辅以目标频带能量比判别车削颤振的持续发展状态,探明切削颤振演化过程与信号特征之间的映射关系。试验结果表明,该监测方法能够准确地判别出颤振从稳定到萌生再到成熟的演变过程,可为后续的颤振控制动作增加响应时间。更多还原

【Abstract】 Machining chatter is a major factor restricting machining quality and efficiency. Traditional chatter monitoring methods have shortcomings in rapid identification of early chatter. In this study, a wavelet packet energy kurtosis index using cutting sound pressure signals was proposed to monitor the early turning chatter. The vibration characteristics of the cutting process from stable to unstable state were analyzed. Wavelet packet was used to decompose the on-line sound pressure signal, obtaining the weak features of the early chatter. Based on wavelet packet energy of each node, the wavelet packet energy kurtosis index was constructed to characterize the incubation period of chatter, and the target frequency band energy ratio was introduced to identify the developing state of chatter. The mapping relationship between the chatter evolution and the signal characteristics has been determined. Finally, the experimental results show that the monitoring method enables identifying the chatter evolution process accurately, which will help to increase the response time for the subsequent chatter control actions.更多还原