基于改进AHP-FCE法的绿色房地产项目评估

为提高往复泵诊断的速度和精度,提出一种通过小波阈值分析处理往复泵振动信号的故障诊断方法。通过小波阈值分析,可以有效去除往复泵振动信号与故障无关的振动信息,然后进一步提取振动信号的归一化能量,将其作为特征值。将特征值与小波包能量分解图综合分析,结合概率神经网络(PNN)对采集后的信号进行往复泵泵阀故障模式进行识别。实验结果表明:小波阈值分析与PNN结合,可以将往复泵泵阀故障类型准确识别,提高了诊断的效率,可以为工业上往复泵的使用和维修大大节约成本,也为往复机械的故障诊断提出了新的解决思路。

Fault Diagnosis of Reciprocating Pump Vibration Signal Based on WaveletAnalysis

To increase the speed and accuracy of fault diagnosis of reciprocating pumps, we propose an approach to the treatment of the vibration signals through wavelet threshold analysis. Through this analysis, we can effectively rid of fault-irrelevant vibration information of the reciprocating pump signals, and then further extract the normalized energy of the vibration signals and use it as the eigenvalue. We make an integrated analysis of the eigenvalue and the wavelet packet energy decomposition diagram, and use PNN to identify the fault modes of the reciprocating pump valve through collected and processed signals. Through experimental analysis, we conclude that wavelet threshold analysis combined with PNN pattern recognition can greatly improve the efficiency of fault diagnosis and save the cost of use and maintenance of reciprocating pumps.