基于粒子群优化的油井液位测试多算法融合研究及其应用

Research and application of oil well liquid level measurement method using multi-algorithm combination based on particle swarm optimization

针对目前油田螺杆泵井动液位深度测量不准确的问题,提出了采用多算法融合构建油井动液位测深模型:利用粒子群优化算法(Particle Swarm Optimization,简称PSO)对3种声速测试算法进行融合,提高了声速的测量精度;通过多特征推理方法避免伪液位回波的干扰,精确辨识真实的液位回波。分别对半实物仿真平台和某油田螺杆泵井现场采集到的数据进行测试,实验结果验证了多算法融合方法对提高油井动液位测量精度的有效性。

In order to accurately measure the dynamic fluid level of oil well,a dynamic oil well fluid level measurement model is established by multi-algorithm combination: three sound velocity test algorithms are combined using particle swarm optimization algorithm to improve the measurement accuracy of sound velocity; the real liquid level echo is accurately identified by multi-features reasoning to avoid the interference of pseudo liquid level echo. The data collected from semi-physical simulation platform and an oilfield screw pump well verify the effectiveness of the multi-algorithm combination oil well liquid level measurement method.