农田水利设备中抽水泵机械振动故障检测方法

针对抽水泵机械振动故障检测中存在的检测效果差的问题,提出一种新的农田水利设备中抽水泵机械振动故障检测方法。分析农田水利设备中抽水泵机械振动信号,提取任意时段特征频率,采用小波包多层分解滤除噪声。对不同数组统计和分析,通过多尺度模糊熵提取抽水泵机械振动故障特征;优化选择测点和特征向量,结合最小二乘支持向量机( LSSVM ),采用多层结构进行多次特征学习,区分故障特征,实现抽水泵机械振动故障检测。实验结果表明,采用所提方法可准确提取抽水泵机械振动故障信号,并且能够准确区分正常信号和故障信号,验证了该方法的可行性。

Detection Method of Mechanical Vibration Fault of Pumping Pump in Farmland Water Conservancy Equipment

Aiming at the problem of poor detection effect in mechanical vibration fault detection of pumping pump , a new method for mechanical vibration fault detection of pumping pump in farmland water conservancy equipment is proposed.The mechanical vibration signal of pumping pump in farmland water conservancy equipment is analyzed , the characteristic frequency in any period is extracted , and the noise is filtered by wavelet packet multi-layer decomposition.According to the statistics and analysis of different arrays , the fault characteristics of mechanical vibration of pump are extracted by multi-scale fuzzy entropy ; optimal selection of measuring points and feature vectors , combined with least squares support vector machine( LSSVM ), multi layer structure is used for multiple feature learning to distinguish fault features , so as to realize mechanical vibration fault detection of pumping pump.The experimental results show that the proposed method can accurately extract the mechanical vibration fault signal of pumping pump , and can accurately distinguish the normal signal and fault signal , which verifies the feasibility of the method.