应用多种机器学习算法的湍流减阻预测研究  

A Study on Predicting Turbulence Drag Reduction Using Multiple Machine Learning Algorithms

作　　者：刘雪婷 梁月 张昕 王倓 LIU Xueting;LIANG Yue;ZHANG Xin;WANG Tan(College of Big Data and Basic Sciences,Shandong Institute of Petrochemical Technology,Dongying 257061,China;College of Petroleum Engineering,Shandong Institute of Petrochemical Technology,Dongying 257061,China)

机构地区：[1]山东石油化工学院大数据与基础科学学院,山东东营257061 [2]山东石油化工学院石油工程学院,山东东营257061

出　　处：《山东化工》2024年第21期57-59,共3页Shandong Chemical Industry

基　　金：山东石油化工学院大学生创新创业训练计划项目(202313386138)。

摘　　要：湍流减阻在工程应用中具有重要意义,能够显著提高流体系统的效率和性能。本文提出了基于三种机器学习算法的方法来预测湍流减阻效果:最小二乘支持向量回归(LS-SVR)、多层感知器(MLP)和粒子群优化(PSO)后的MLP。将收集到的湍流数据提取关键特征后,分别训练和优化这三种算法,比较它们在湍流减阻预测中的性能。其中,PSO优化后的MLP在预测精度和计算效率方面表现最佳。本文的研究为利用机器学习技术优化湍流减阻提供了新的见解与方法。Turbulence drag reduction holds significant importance in engineering applications,as it can markedly enhance the efficiency and performance of fluid systems.This paper proposes a method based on three machine learning algorithms to predict the effects of turbulence drag reduction:Multilayer Perceptron(MLP),Particle Swarm Optimization(PSO)enhanced MLP,and Least Squares Support Vector Regression(LS-SVR).Firstly,turbulence data was collected and processed to extract key features.Subsequently,the three algorithms were trained and optimized separately,and their performance in predicting turbulence drag reduction was compared.Experimental results indicate that the PSO-optimized MLP outperforms the other methods in terms of prediction accuracy and computational efficiency.This study provides new insights and methods for optimizing turbulence drag reduction using machine learning techniques.

关 键 词：粒子群优化 MLP LS-SVR 机器学习 湍流减阻 预测 

分 类 号：TQ015.9[化学工程]