数控机床机械局部故障智能检测方法仿真

Simulation of Intelligent Detection Method for Local Faults of CNC Machine Tools

研究一种有效的故障智能检测方法,能够减少运行时间,降低检测能耗及误差率,在实际应用中具备一定的应用价值。为了解决当前检测方法因局部故障比较复杂导致局部故障检测能耗过高,检测误差较大,运行时间较长等问题,提出一种基于局部均值分解的数控机床机械局部故障智能检测方法,通过收集机械局部故障振动信号,利用小包分解法对故障振动信号进行去噪,基于新的阈值去噪量化法重构数控机床机械局部故障信号。利用局部故障信号计算相邻极值的包络估计值,利用滑动平均法获取调频信号,将调频信号作为初始信号输入,多次迭代后,利用能量算子解调法对初始信号分量进行处理,输出局部故障信号的包络谱,完成数控机床机械局部故障检测。仿真结果表明,所提方法可缩短检测的运行时间,降低检测能耗及误差,在实际应用中具有一定的应用价值。

An effective method to detect the fault intelligently can reduce running time, detection energy consumption and error rate. Therefore, an intelligent detection method for local mechanical fault of numerical control machine tools based on local mean decomposition was proposed. After collecting the mechanical local fault vibration signal, the packet decomposition method was used to remove the noise in fault vibration signal. Based on new threshold denoising quantization method, the local fault signal of numerical control machine tools was reconstructed. Moreover, local fault signal was used to calculate the envelope estimation value of adjacent information detection. Meanwhile, sliding mean method was used to obtain FM signal, and then FM signal was used as the initial signal input. After much iteration, the energy operator demodulation method was applied to the initial signal component, and the envelope spectrum of local fault signal was output. Thus, the mechanical failure detection of numerical control machine tools was completed. Simulation results show that the proposed method can shorten the running time of detection and reduce the energy consumption and error of detection, which has certain application value.