摘要
悬点示功图是判断抽油机井生产状况的有效手段之一,针对目前电参转功图的难点问题,提出了一种基于FOA-BP神经网络的抽油机悬点示功图反演方法。采集测试抽油机电功率并将其转化为光杆功率,将抽油机光杆功率和扭矩因数作为网络的输入参数,通过模型训练消除抽油机结构及平衡参数对示功图反演计算的影响,再经过小波变换去噪处理,完成由电动机功率向悬点示功图的反演计算。对现场160井次实例计算表明,反演示功图与实测示功图吻合度达95.18%。基于FOA-BP神经网络的抽油机悬点示功图反演技术为及时、准确判断抽油机系统井下运行状态提供了理论技术支撑。
The suspension point indicator diagram is one of the effective means to judge the production situation of pumping wells.In response to the current difficulties in converting electrical parameters to power diagrams,a FOA-BP neural network-based inversion method for the suspension point indicator diagram of pumping units is proposed.The electric power of the test pumping unit is collected and converted into polished rod power.The polished rod power and torque factor of the pumping unit are used as input parameters for the network.The influence of the pumping unit structure and balance parameters on the inversion calculation of the indicator diagram is eliminated through model training.After wavelet transform denoising processing,the inversion calculation from the motor power to the suspension point indicator diagram is completed.The calculation of 160 well instances on site shows that the consistency between the inversion indicator diagram and the measured indicator diagram is 95.18%.The inversion technology of the suspension point indicator diagram of the pumping unit proposed based on the FOA-BP neural network provides theoretical and technical support for timely and accurate judgment of the underground operation status of the pumping unit system.
作者
卢玉
LU Yu(Drilling & Production Technology Research Institute of Liaohe Oilfield Company)
出处
《油气田地面工程》
2023年第10期58-63,共6页
Oil-Gas Field Surface Engineering
关键词
悬点示功图
电动机功率
扭矩因数
光杆功率
FOA-BP算法
suspension point indicator diagram
motor power
torque factor
polished rod power
FOA-BP algorithm
