基于AO-ELM算法的旋风分离器粒级效率建模研究  

Research on Particle Size Efficiency Modeling of Cyclone Separator Based on AO-ELM Algorithm

作　　者：尚栋栋 李绪明 卢鹏飞 陈卓 SHANG Dongdong;LI Xuming;LU Pengfei;CHEN Zhuo(Shaanxi Gas Group Transportation Energy Development Co.,Ltd.;Desert Transportation Branch of China Petroleum Transportation Co.,Ltd.;Changqing Engineering Design Co.,Ltd.,CNPC;No.2 Oil Production Plant of Northwest Oilfield Company,SINOPEC)

机构地区：[1]陕西燃气集团交通能源发展有限公司,陕西省西安市710016 [2]中国石油运输有限公司沙漠运输分公司 [3]中国石油长庆工程设计有限公司 [4]中国石化西北油田分公司采油二厂

出　　处：《油气田地面工程》2023年第8期62-69,共8页Oil-Gas Field Surface Engineering

摘　　要：旋风分离器是气田生产过程中的重要设备,其主要作用是将采出气中含带的固体砂粒进行旋流分离,然而目前针对旋风分离器分离性能的研究主要停留在实验室阶段。为了较为准确地预测大尺寸旋风分离器粒级效率,首次使用天鹰优化器(Aquila Optimizer)优化极限学习机(Extreme Learning Machine)建立了AO-ELM旋风分离器粒级效率预测算法模型。通过与所建其他模型预测数值对比得出:AO-ELM算法模型相较与AO-BPNN、AO-ENN模型预测精度更高,预测数值更接近于实测值;相较于旋风分离器分离效率理论和半经验模型同样具有更精确的数值预测,对旋风分离器工业设计具有一定的指导意义。Cyclone separator is an important equipment in the production process of gas field.Its main function is to carry out cyclone separation of solid sand particles in the produced gas.However,at pres-ent,the research on the separation performance of cyclone separators mainly stays in the laboratory stage.In order to accurately predict the particle size efficiency of large-size cyclone separators,the AO ELM cyclone separator particle size efficiency prediction algorithm model is established by using Aquila Optimizer to optimize Extreme Learning Machine for the first time.Compared with other models,the results show that the prediction accuracy of the AO-ELM algorithm model is higher than that of the AO-BPNN and AO-ENN models,and the prediction value is closer to the measured value.Com-pared with cyclone separator separation efficiency theory and semi-empirical model,it also has more accurate prediction values,which has certain guiding significance for cyclone separator industrial design.

关 键 词：天鹰优化器 旋风分离器 极限学习机 预测模型 粒级效率 

分 类 号：TE96[石油与天然气工程—石油机械设备]