中文摘要：

为研究数控机床丝杠副性能退化机理,对丝杠副性能进行评估。首先采用小波包对丝杠副螺母座、轴承座的振动信号进行分解,提取小波包分解后的各阶功率谱作为特征参数,分析丝杠进给速度、切削深度对丝杠副振动特性的影响。利用BP神经网络建立丝杠副性能退化评估模型。通过振动信号、电机驱动电流信号、进给速度、切削深度以及加工方案等评估丝杠副性能退化状态,实验证明该性能退化评估模型准确率较高。

英文摘要：

Analysis vibration signals of bearing seat and nut seat with wavelet packet. Then take power spectrum of wavelet packet to be characteristic parameters. By these parameters, study performance degradation of screw pair for NC machine tool. Compare power spectrum of vibration in different feed rates and cutting depths, and try to find the effect of feed rate and cutting depth on screw vibration. In order to evaluate performance of screw pair, build performance degradation model based on BP neural network. The input parameters include vibration, current sig- nals of screw motor, feed rate, cutting depth and processing scheme, and the output of evaulation model is the re- sult of performance degradion for screw pair. Tests show that the evauation results are important reference for main- tance of scrwe pair.