采用智能进化算法的管壳式换热器详细设计  

Detailed design of shell-and-tube heat exchanger using intelligent evolutionary algorithms

作　　者：李海东 张奇琪 杨路 AKRAM Naeem 常承林 莫文龙[3,4] 申威峰 LI Haidong;ZHANG Qiqi;YANG Lu;AKRAM Naeem;CHANG Chenglin;MO Wenlong;SHEN Weifeng(National-municipal Joint Engineering Laboratory for Chemical Process Intensification and Reaction,School of Chemistry and Chemical Engineering,Chongqing University,Chongqing 401331,China;School of Chemical Engineering,Minhaj University Lahore,Lahore 54000,Punjab,Pakistan;State Key Laboratory of Carbon-based Energy Resource Chemistry and Utilization,Xinjiang Key Laboratory of Clean Coal Transformation and Chemical Process,College of Chemical Engineering,Xinjiang University,Urumqi 830017,Xinjiang,China;Xinjiang Zhundong Coal High Value Diversified Engineering Technology Research Center,Xinjiang Yihua Chemical Company Limited,Changji 831100,Xinjiang,China)

机构地区：[1]重庆大学化学化工学院,化工过程强化与反应国家地方联合工程实验室,重庆401331 [2]明哈吉大学拉合尔分校化工学院,巴基斯坦旁遮普拉合尔54000 [3]新疆大学化工学院,省部共建碳基能源资源化学与利用国家重点实验室,煤炭清洁转化与化工过程新疆维吾尔自治区重点实验室,新疆乌鲁木齐830017 [4]新疆宜化化工有限公司,新疆准东煤高值多元化工程技术研究中心,新疆昌吉831100

出　　处：《化工学报》2025年第1期241-255,共15页CIESC Journal

基　　金：新疆维吾尔自治区区域协同创新专项科技援疆计划项目(2024E02036);中央高校基本科研业务费专项资金项目(2024CDJXY010);重庆市出站留(来)渝博士后择优资助项目(Z20240373);重庆市技术创新与应用发展重点专项项目(2024TIAD-KPX0168);国家自然科学基金项目(22008210)。

摘　　要：管壳式换热器是石油、化工等过程工业中应用最广泛的热量回收设备,其数学模型通常是十分复杂的非线性优化问题,现有的商业求解器和优化算法存在运算时间长、收敛困难、易陷入局部最优等难题。针对这些难题,参考管壳式换热器制造标准,将换热器内构件尺寸定义成离散变量,分别以最小化换热面积、年度总费用、环境影响因子及最大化传热效率为目标函数,建立管壳式换热器详细设计的混合整数非线性规划模型。同时,对传统智能进化算法包括遗传算法、粒子群算法及模拟退火算法进行改进,使得换热器设计变量能够在一系列离散值中自由选择,不需要对优化结果进行人工圆整处理。案例测试结果表明,改进的智能进化算法能在1.0 s内得到最优设计方案,相对于全局求解器,优化时间节约99%以上,提高了优化求解效率;相对于局部求解器,改进的智能进化算法能够获取全局最优解,换热面积节约15.4%~56.6%,年度总费用节约15.8%~77.8%,保证设计质量。通过多目标优化在不同目标函数之间进行权衡,通过灵敏度分析展示了不同设计变量对目标函数的影响趋势。The shell heat exchanger is the most widely used calorie recovery equipment in the process of oil,chemical industry,etc.,and its mathematical models are usually very complicated non-linear optimization problems.The existing commercial solution device and optimization algorithm have long computing time,and the computing time is long.It is difficult to converge and easily fall into part of the optimal problem.In order to address these challenges,this research draws inspiration from the manufacturing norms of shell-and-tube heat exchangers.It delineates the dimensions of the internal components of heat exchangers as discrete variables and formulates a mixed integer nonlinear programming model for the intricate design of shell-and-tube heat exchangers.The model aims to minimize heat transfer area,annual total cost,environmental impact factor,and maximize heat transfer efficiency.At the same time,enhancements have been made to the traditional intelligent evolutionary algorithm,which includes the genetic algorithm,particle swarm optimization algorithm,and simulated annealing algorithm.This allows for the selection of design variables for the heat exchanger from a range of discrete values,eliminating the need for manual rounding of optimization results.Test results demonstrate that the enhanced intelligent evolutionary algorithm can achieve the optimal design solution within 1.0 s,reducing optimization time by over 99%compared to the global solver and enhancing optimization solution efficiency.Compared to the local solver,the enhanced intelligent evolution algorithm can achieve the global optimal solution,reduce heat transfer area by 15.4%—56.6%,lower annual total cost by 15.8%—77.8%,and guarantee design quality.Furthermore,multiobjective optimization is utilized to balance different objective functions,while sensitivity analysis results illustrate the impact of various design variables on the objective functions.

关 键 词：传热 管壳式换热器 优化设计 智能进化算法 多目标优化 

分 类 号：TQ051.5[化学工程]