基于改进变分模态分解的抽油井偏磨程度诊断

Diagnosis of the Degree of Wear of Oil Wells Based on Improved Variational Mode Decomposition

目前抽油井工况分析方法与实时智能诊断技术不完善,无法及时发现、处理偏磨问题,导致抽油杆、泵等关键部件存在严重的损坏风险。为此,提出一种基于改进变分模态分解(IVMD)的抽油井偏磨程度诊断方法,其核心思想在于,扭矩和轴向力的变化会导致抽油井的偏磨程度发生改变,从而影响电参数信号的频率和幅值。首先通过改进人工鱼群算法优化变分模态分解(VMD)的分解层数与惩罚因子,然后将油井电参数信号分解成多个局部振动模态,并对生成的各局部振动模态进行特征分析,最后采用RGB图实现对抽油井偏磨程度诊断。研究结果表明,该方法可有效判断偏磨程度。

The current methods for analyzing the working condition of pumping wells and diagnosing issues in real-time are not yet flawless.This results in the inability to detect and address eccentric wear problems in a timely manner,which places crucial components like the sucker rod and pump at a significant risk of damage.Therefore,a diagnostic method for the degree of pump wear in pumping wells based on improved variational mode decomposition(IVMD)is proposed.The central idea is that changes in torque and axial force can cause variations in the degree of pump wear,thereby affecting the frequency and amplitude of electrical parameter signals.Firstly,the number of decomposition layers and penalty factors of variational mode decomposition(VMD)is optimized by improving the artificial fish swarm algorithm.Subsequently,the electrical parameters of the oil well are decomposed into multiple local vibration modes and perform characteristic analysis on the generated local vibration modes.Finally,the RGB image is employed for diagnosing the degree of eccentric wear in the pumping unit.The research results show that this method can be used to effectively judge the severity of eccentric wear.