中文摘要：

为减少数控机床的维护维修成本,减少机床停机时间,依据小波神经网络建立数控机床部件的故障预测模型,并对振动信号、电流信号以及速度信号进行处理,提取各信号的特征值训练机床部件故障预测模型.通过训练后的故障预测模型对滚珠丝杠副、导轨副以及主轴系统等进行故障预测,综合部件故障预测结果和机床传统维修方案制定最优的维修方案、维护计划和备件计划,避免因备件数量不当、维修不当及保养不佳造成的机床维护维修成本.智能维护系统有效保证了数控机床的加工精度并提高生产效率.

英文摘要：

In order to reduce the maintenance cost and the downtime of NC machine tool, it built fault prediction model of NC machine tool parts based on wavelet neural network. Then dealt with vibration signal, current signal and velocity signal. It took eigenvalue to train the fanlt prediction model of NC machine tool. Predicted fault of the parts of NC machine tool, such as ball screw, spindle and so on. Based on the results of prediction and traditional maintenance plan it built the optimal maintenance plan and spare parts scheme. Thus, avoided the maintenance cost of NC machine tool because of unfit quantity of spare parts and incorrect maintenance. The intelligent maintenance system is used to guarantee machining precision and production efficiency.