原油管道蜡沉积速率预测模型研究  

Prediction model of wax deposition rate in crude oil pipeline

作　　者：肖荣鸽 刘博 刘国庆 戴政 张鹏 XIAO Rong-ge;LIU Bo;LIU Guo-qing;DAI Zheng;ZHANG Peng(Shaanxi Key Laboratory of Advanced Stimulation Technology for Oil&Gas Reservoirs,Xi′an Shiyou University,Xi′an 710065,Shaanxi Province,China;Maintenance and Emergency Repair Branch of Xi′an Northwest Petroleum Pipeline Co.,Ltd.,Xi′an 710016,Shaanxi Province,China)

机构地区：[1]西安石油大学,陕西省油气田特种增产技术重点实验室,陕西西安710065 [2]西安西北石油管道有限公司维抢修分公司,陕西西安710016

出　　处：《化学工程》2023年第3期78-83,共6页Chemical Engineering(China)

基　　金：陕西省教育厅2019年度服务地方专项计划项目(19JC034);陕西省科技厅重点研发计划项目(2021GY-139)。

摘　　要：准确地预测原油管道蜡沉积速率能够有效确定清管周期,以保证原油管道安全运行。针对BP神经网络(BPNN)模型学习效率低、对初始权重敏感且容易陷入局部最优状态等缺点,采用改进的麻雀搜索算法(ISSA)来优化BPNN的初始权值和阈值,建立ISSA-BPNN蜡沉积速率预测模型。以华池作业区38组蜡沉积实验数据为研究对象,使用MATLAB软件搭建预测模型并进行预测,同时与BPNN模型、遗传算法优化的BPNN模型(GA-BPNN)、粒子群优化算法优化的BPNN模型(PSO-BPNN)以及SSA-BPNN模型进行对比分析。结果表明:ISSA-BPNN模型预测蜡沉积速率的平均相对误差为1.3531%,决定系数R^(2)为0.9948,均优于BPNN、GA-BPNN、PSO-BPNN和SSA-BPNN模型的预测结果,证明了ISSA-BPNN模型作为预测管道蜡沉积速率工具的准确性和可行性。Accurately predicting the wax deposition rate of crude oil pipelines can determine the pigging cycle effectively and ensure the safe operation of crude oil pipelines.Aiming at the shortcomings of BP neural network(BPNN)model,such as low learning efficiency,sensitive to initial weights and easy to fall into local optimal state,an improved sparrow search algorithm(ISSA)was used to optimize the initial weights and thresholds of BPNN,and ISSA-BPNN wax deposition rate prediction model was established.Taking 38 sets of experimental data of wax deposition in Huachi operation area as the research object,the prediction model was built and predicted using MATLAB software.At the same time,it was compared with BPNN model,BPNN model optimized by genetic algorithm(GA-BPNN),BPNN model optimized by particle swarm optimization algorithm(PSO-BPNN)and SSA-BPNN model.The results show that the mean absolute percentage error of ISSA-BPNN model in predicting the wax deposition rate is 1.3531%,and the coefficient of determination R^(2) is 0.9948,which are better than the prediction results of BPNN,GA-BPNN,PSO-BPNN and SSA-BPNN models.Therefore,ISSA-BPNN model has high accuracy and feasibility,and can be used to predict the wax deposition rate of crude oil pipelines.

关 键 词：蜡沉积速率 含蜡原油 麻雀搜索算法 BP神经网络 Tent混沌映射 

分 类 号：TE832[石油与天然气工程—油气储运工程]