基于神经网络和正交实验设计的CO_(2)柱塞泵容积效率的数值模拟分析  

Numerical Simulation Analysis of Volumetric Efficiency of CO_(2) Plunger Pump Based on Neural Network and Orthogonal Experimental Design

作　　者：王喜印 周三平[1] WANG Xiyin;ZHOU Sanping(Institute of Mechanical Engineering,Xi'an Shiyou University,Xi'an 710065,China)

机构地区：[1]西安石油大学机械工程学院,陕西西安710065

出　　处：《化工技术与开发》2024年第12期70-76,共7页Technology & Development of Chemical Industry

摘　　要：针对CO_(2)柱塞泵的效率低下的问题,本文通过数值模拟方法和正交实验设计,对柱塞泵的7个几何尺寸对柱塞泵容积效率的影响进行了数值模拟,并进行了极差分析和方差分析。结果表明,柱塞直径、柱塞分布圆直径、配流盘腰形槽宽度、斜盘倾角和配流盘腰形槽张角对容积效率的影响依次降低,是显著性影响因素,其中容积效率随柱塞直径、分布圆直径和斜盘倾角的增大而增大,随配流盘腰形槽宽度的增大而直线下降。小于140°时,腰形槽的张角对容积效率几乎无影响,大于140°以后,容积效率快速下降。进出口直径的影响可忽略不计。采用神经网络方法,构建了预测柱塞泵容积效率的神经网络模型,预测值与数值模拟值的最大误差不超过2.48%,可为二氧化碳柱塞泵的设计及容积效率预测提供参考。Addressing the issue of low efficiency in CO_(2) piston pumps,the influence of seven geometric dimensions of the plunger pump on the volumetric efficiency of the plunger pump was numerically simulated through numerical simulation methods and orthogonal experimental design,the range analysis and variance analysis were carried out.The results indicated that the effects of plunger diameter,plunger distribution circle diameter,port plate waist groove width,inclined plate angle,and port plate waist groove opening angle on volumetric efficiency decreased sequentially,and were significant influence factors.Moreover,volumetric efficiency increased with the increase of plunger diameter,distribution circle diameter,and inclined plate angle,and decreased linearly with the increase of port plate waist groove width.When it was less than 140°,the waist groove opening angle had almost no effect on the volumetric efficiency,and after it was greater than 140°,the volumetric efficiency sharply decreased.The impact of import and export diameter could be ignored.Using neural network methods,a neural network model was constructed to predict the volumetric efficiency of a plunger pump.The maximum error between the predicted value and the numerical simulation value did not exceed 2.48%,which could provide guidance for the design and volumetric efficiency prediction of carbon dioxide plunger pumps.

关 键 词：CO_(2)柱塞泵 神经网络 正交试验设计 容积效率 数值模拟 

分 类 号：TQ051.21[化学工程]