复杂工况下油田机械设备运行状态监测方法

Monitoring Method for Operating State of Oilfield Machinery under Complex Working Conditions

提出一种复杂工况下的油田机械设备运行状态监测方法,采用EMD方法对油田机械设备的振动信号进行去噪处理,结合ITD算法提取油田机械设备振动信号幅频特征,输入Teager能量算子获得振动信号幅频特征,运用SOFM网络分析该信号的幅频特性,得到特征聚类结果,在此基础上建立二叉树支持向量机,将特征聚类结果输入进去,完成油田机械设备运行状态的监测识别。实验结果表明,所提方法的监测性能良好,具有较高的监测效率。

A monitoring method for the operation status of oilfield machinery under complex working conditions was proposed,including having EMD method adopted to denoise vibration signals of oilfield machinery and the ITD algorithm combined to extract amplitude frequency characteristics of their vibration signals,and the Teager energy operator input to obtain amplitude frequency characteristics of the vibration signals as well as the SOFM network employed to analyze amplitude frequency characteristics of the signals so as to get feature clustering results.On this basis,the binary tree support vector machine was established and the feature clustering results were input to complete monitoring and identification of operating status.The experimental results show that,the method proposed has better monitoring performance and high efficiency.